AI-BASED MEDICAL QnA SERVICE METHOD AND DEVICE

Abstract

The method includes the steps of receiving user input data including user query data from a user terminal, generating AI-based response data using at least one of previously stored question-response data and medical academic data as a response to the user input data, and providing the response data to the user through the user terminal.

Description

Technology Field

[0001] The present invention relates to a medical Q&A service utilizing artificial intelligence (AI), and to an AI-based medical Q&A service method and apparatus that analyzes user query data and provides responses based on medical data and academic information. Background Technology

[0002] The following description merely provides background information related to an embodiment according to the present invention and does not constitute prior art.

[0003] Existing medical consultation methods suffer from limited accessibility for many users due to time and spatial constraints as well as cost issues. While hospital visits or face-to-face consultations with medical professionals can provide accurate information, there are limitations in medical resources that prevent the provision of the same service to all users.

[0004] Furthermore, users struggle to obtain accurate medical information because searching for it via the internet increases the risk of exposure to misinformation or unreliable data. In particular, since medical information requires complexity and expertise, there is a problem in that it is difficult for general users to understand and apply it to their own situations. Existing automated Q&A systems are limited to providing simple information or responding only to standardized query formats, showing limitations in processing diverse user inputs and generating customized responses tailored to individual circumstances. Along with this, there is also the issue that users find it difficult to evaluate the reliability of responses because it is not clear what data the generated responses are based on.

[0005] Recently, AI-based medical Q&A methods are emerging that provide medical information in a user-friendly manner while processing various input formats and ensuring the reliability of responses. Prior art literature

[0006] Republic of Korea Published Patent Application No. 10-2023-0111480 The problem to be solved

[0007] Existing medical Q&A services suffer from issues such as an inability to effectively provide users with the information they desire, or a lack of information reliability and contextual relevance. Face-to-face consultations are time-consuming and costly, while existing automated systems have limitations in processing various input formats or providing personalized responses. In particular, it is difficult to secure user trust because the basis for response data is not clearly presented, and there is a lack of functionality to efficiently manage or share generated information.

[0008] Furthermore, data formats such as posts fail to adequately reflect the context and timing of the conversation, which degrades the relevance and usefulness of responses. This issue also leaves something to be desired in striking a balance between protecting user privacy and sharing information.

[0009] Recently, there is a demand for reliable AI-based medical Q&A services that accommodate various data formats, provide customized responses considering context and timing, clearly present evidence, and feature flexible information management capabilities. means of solving the problem

[0010] The method may include the steps of receiving user input data including user query data from a user terminal, generating AI-based response data using at least one of previously stored query response data and medical academic data as a response to the user input data, and providing the response data to the user through the user terminal.

[0011] The above query data is in the form of a message entered through a chat window or a post posted as the start post of a thread, and the step of providing the above response data may include the step of providing the above response data as a response message or a comment of the above thread in correspondence with the format of the above query data.

[0012] The step of providing the above response data may include the step of providing the above response message at a first time point if the above query data is in the message format, or the step of providing the above comment at a second time point later than the first time point if the above query data is in the post format.

[0013] The step of providing the above response data may include, if the above query data is in the form of the above post and another user's comment is posted in the above thread before the above second time point, the step of providing the comment at a third time point later than the above second time point.

[0014] The step of generating the above response data may include the step of updating the above response data based on the comment if the above query data is in the form of the above post and a comment from another user is posted in the above thread before the above second time point.

[0015] The step of providing the above response data may include the step of providing the response data in a private state so as to be exposed only to the user, and the step of switching the response data to a public state so as to be exposed to other users according to the user's choice.

[0016] The step of switching to the public state may include, when the query data is in the form of the post and the user adopts the response data, the step of switching the response data from the private state to the public state.

[0017] The step of providing the above response data may include the step of mapping at least one of the question-answer data and medical academic data used to generate a part of the above response data as reference data, and the step of providing the above reference data together with the above response data.

[0018] The step of providing the above response data may include the step of displaying a first object so as to be associated with a part of the mapped response data of the reference data, and, if there is conflicting content between the mapped reference data, the step of displaying a second object distinguished from the first object so as to be associated with a part of the mapped response data of the conflicting reference data.

[0019] The system includes a memory on which at least one program is recorded and a processor that executes said program, and said program may include instructions for performing the steps of: receiving user input data including user query data from a user terminal; generating artificial intelligence-based response data using at least one of previously stored query response data and medical academic data as a response to said user input data; and providing said response data to the user through said user terminal. Effects of the invention

[0020] According to the present invention, user experience can be improved by analyzing query data of various formats and providing customized responses based on reliable medical information.

[0021] Furthermore, it clearly presents the basis for responses and simultaneously supports privacy protection and information sharing through a function to switch between private and public states. This effectively enhances the accessibility and reliability of medical information. Brief explanation of the drawing

[0022] FIG. 1 shows subjects illustrating an AI-based medical Q&A service method according to one embodiment of the present invention. FIG. 2 is a flowchart illustrating an AI-based medical Q&A service method according to one embodiment of the present invention. FIG. 3 is a flowchart illustrating an AI-based medical Q&A service method according to another embodiment of the present invention. FIG. 4 is an example of a screen showing the execution of an AI-based medical Q&A service according to one embodiment of the present invention. FIG. 5 is an example of a screen showing an AI-based medical Q&A service executed according to one embodiment of the present invention. FIG. 6 is an example of another screen showing an AI-based medical Q&A service executed according to one embodiment of the present invention. FIG. 7 is a diagram showing a computer system according to one embodiment of the present invention. Specific details for implementing the invention

[0023] The present invention will be described in detail below with reference to the accompanying drawings. Hereinafter, repetitive descriptions and detailed descriptions of known functions and configurations that may unnecessarily obscure the essence of the invention are omitted. Embodiments of the present invention are provided to more fully explain the invention to those with average knowledge in the art. Accordingly, the shapes and sizes of elements in the drawings may be exaggerated for clearer explanation.

[0024] Throughout the specification, when a part is described as "comprising" a certain component, this means that, unless specifically stated otherwise, it does not exclude other components but may include additional components.

[0025] Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the attached drawings.

[0026] FIG. 1 shows subjects illustrating an AI-based medical Q&A service method according to one embodiment of the present invention.

[0027] Referring to FIG. 1, an AI-based medical Q&A service method according to one embodiment of the present invention can be implemented with a user terminal (110), a server (120), and a database (130).

[0028] Specifically, the user terminal (110) may be a primary interface device that supports the user in inputting query data, checking generated response data, and exploring or managing additional information as needed to receive AI-based medical Q&A services. The user terminal (110) may be implemented as various devices such as smartphones, tablets, laptops, PCs, etc., and may perform AI-based medical Q&A services by connecting with a server (120) and a database (130) via a network.

[0029] Additionally, the user terminal (110) allows the user to naturally input data through a query input interface. The query data can be entered in the form of a message through a chat window or in the form of a post as the start of a thread. In the chat window, data requiring a real-time response or short query data can be entered, while the post format can include longer queries or detailed content. In this way, the user terminal (110) can be designed to process various forms of data without being constrained by the user's query format.

[0030] Input query data is transmitted to the server (120) via the user terminal (110), and AI-based response data generated by the server (120) can be returned to the user terminal (110) and displayed. The user terminal (110) can support an intuitive output screen to provide the generated response data in a user-friendly manner. For example, a response in message format is immediately displayed in a chat window, and a response in post format is provided in the form of a comment, and in this process, the basis of the response data can be displayed together in the form of annotations.

[0031] Furthermore, the user terminal (110) can provide various functions to explore and manage the generated response data. The user can quickly navigate to detailed areas of the response data or academic evidence materials through exploration buttons such as the Moreden button or the KCI button. Such exploration functions support the efficient use of information and allow for the systematic verification of complex medical information.

[0032] Switching between private and public states may also be one of the main functions of the user terminal (110). The user can view response data in a private state and, if necessary, switch it to a public state to share information with other users. For example, if the user accepts the provided response data, the data is switched from a private state to a public state, and other users can also view it. This function can enhance user convenience while maintaining a balance between information sharing and privacy protection.

[0033] The server (120) can perform the core role of receiving query data from the user terminal (110), processing it to generate AI-based response data, and then providing it to the user terminal (110).

[0034] The server (120) can support the overall functions of the service by connecting with the database (130) to search for data related to user queries, generating reliable information based on this, and managing the state of response data.

[0035] Specifically, the server (120) can first receive and analyze query data. Query data in the form of messages or posts transmitted from the user terminal (110) is analyzed by the server (120) using natural language processing (NLP) technology, and the intent of the query and key keywords can be identified to lay the foundation for generating an appropriate response. For example, when a question about a specific symptom is entered, the server can analyze the context of the symptom and extract the core content of the query to prepare for data retrieval and response generation.

[0036] Next, the server (120) performs response data generation and can generate response data that matches the user's query by utilizing existing query response data and medical academic data stored in the database (130). During the generation process, the server (120) selects data highly relevant to the user's query through an AI algorithm and can construct reliable information based on this. In the case of a query in the form of a post, it can generate context-appropriate information by analyzing comments from other users within the thread to update response data or provide a supplemented response.

[0037] The generated response data can be provided in an appropriate manner depending on the format of the user's query. For message formats, it is provided as a real-time response message, and for post formats, it is provided as a comment; in this process, the basis for the response is displayed in the form of annotations, allowing the user to directly verify the reliability of the information.

[0038] Additionally, the server (120) can flexibly control the state of the response data through private and public state management. Initially, the response data is provided in a private state, but it can be switched to a public state upon a user's request or when specific conditions are met. For example, if a user accepts the response data, the server (120) can automatically switch it to a public state to provide the information to other users as well. This feature can enhance the reliability of the service while maintaining a balance between protecting user privacy and sharing information.

[0039] Finally, the server (120) can continuously improve the performance of the service through data learning and quality improvement. The server (120) can analyze interaction data with users to learn recurring query patterns and generate more sophisticated and personalized response data based on this. This can improve the user experience and enable the provision of an efficient medical Q&A service.

[0040] The database (130) may be a core component that stores and manages question-and-answer data and medical academic data in conjunction with the server (120) and provides data necessary for an AI-based medical Q&A service. The database (130) may be designed to systematically organize various medical information and to quickly search for and provide appropriate data according to user queries.

[0041] The database (130) may first contain previously stored question-and-answer data. This data is organized based on previous user queries and their responses and can be used for training AI algorithms and generating response data. For example, the database (130) may store question-and-answer data related to the symptoms, treatments, and prevention strategies of a specific disease, so that when a similar question is entered, it can be immediately searched to contribute to the generation of response data. Through this, the database (130) can provide quick and consistent responses to questions that users repeatedly request.

[0042] Additionally, the database (130) may include medical academic data to provide the basis for response data that requires reliability and accuracy. The academic data consists of medical papers, guidelines, professional books, or authoritative materials provided by medical institutions and can be used to provide scientific evidence for the content queried by the user. For example, the database (130) may include paper data on the latest treatment methods for specific diseases and add the corresponding evidence in the form of annotations to the response data generated by the AI. This can significantly improve the reliability of the response provided to the user.

[0043] In addition, the database (130) may be equipped with high-performance search algorithms and index structures to efficiently search and manage question-and-answer data and medical academic data. The database (130) can be utilized to generate AI response data by prioritizing the search for the most relevant data based on the user's query data and providing it to the server (120). In particular, the database (130) can quickly provide appropriate data even when the user inputs complex questions, not only through keyword-based search but also through context analysis and similarity-based search.

[0044] In addition, the database (130) can support data updates and scalability. When new question-and-answer data or the latest medical academic data is added, the database (130) can update it to maintain up-to-date information. For example, when a new treatment for a specific disease is announced, the data can be added to the database (130) so that it can be immediately utilized in the process of generating AI response data. This update process allows the database (130) to continuously provide reliable and up-to-date information.

[0045] Furthermore, the database (130) can be structured hierarchically to support various user needs. It can have a structure that allows for the separate management or integrated searching of various data types, such as question-and-answer data, medical academic data, and user interaction data. Through this, the database (130) can satisfy the needs of various user groups, such as medical professionals and general users.

[0046] FIG. 2 is a flowchart illustrating an AI-based medical Q&A service method according to one embodiment of the present invention.

[0047] Referring to FIG. 2, in an AI-based medical Q&A service method according to one embodiment of the present invention, a server (120) can receive user input data including query data through a user terminal (110) (S210). The user input data can be provided in various formats such as messages, posts, and voice input.

[0048] The server (120) can generate AI-based response data in response to user input data by using at least one of previously stored question-and-answer data and medical academic data (S220). In particular, the response data can provide the user with practical answers reflecting the experience of medical professionals and reliable information based on the latest academic materials, based on previously stored question-and-answer data or medical academic data.

[0049] The generated response data is immediately provided through the user terminal (110), allowing the user to quickly obtain clear and appropriate information regarding their question (S230). For example, when a query regarding a specific symptom is entered, the artificial intelligence searches the relevant database (130) to generate a response based on the most suitable information and can deliver it in a way that is easy for the user to understand. This process can provide a personalized medical experience to the user by delivering complex medical information simply and clearly.

[0050] By automatically generating responses based on user input data, artificial intelligence can reduce time and costs during the medical consultation process and significantly improve the user experience. Furthermore, response quality can be enhanced through continuous learning, enabling the provision of high-quality medical information with reliability and accuracy. This lays the foundation for providing medical information to a wider range of users while efficiently utilizing medical resources.

[0051] Therefore, the server (120) can implement a rapid and reliable medical consultation environment through an AI-based Q&A service that provides medical information in a user-friendly and personalized manner.

[0052] As an optional embodiment, the server (120) can improve the user experience by providing response data in a manner appropriate to the format of the query data during the process of processing user input data. The user query data may be provided in the form of a message entered through a chat window or in the form of a post as the start of a thread, and by effectively processing these various data formats, the user can receive information in the manner desired. Since query data in message format often requires an immediate response, and data in post format requires a response that fits the context of the thread, an optimized response can be provided according to each format.

[0053] In particular, for data in the form of messages entered through the chat window, real-time response messages are provided, allowing users to quickly check the necessary information. On the other hand, for data in the form of posts, responses are provided in a comment format tailored to the thread structure, enabling the delivery of appropriate information while maintaining the context of the post. This differentiated response delivery method allows users to conveniently access medical information regardless of the data format and enables the generation of reliable responses even in diverse data environments.

[0054] In addition, the server (120) can achieve real-time performance and contextual relevance simultaneously by innovatively improving the way a user obtains medical information and providing a response optimized according to the format of the data. Through this, the user can receive reliable information in the desired way, whether as a message or a post, and the accessibility and usability of medical information can be greatly improved.

[0055] As an optional embodiment, the server (120) can optimize the user experience and provide context-appropriate information by differentiating the timing of providing response data according to the format of the query data. If the user's query data is in the form of a message entered through a chat window, the response message may be provided at a first time point, considering the characteristic that an immediate response is important. On the other hand, if the user's query data is entered in the form of a post, the response data may be provided in the form of a comment at a second time point later than the first time point to reflect the overall flow and context of the thread.

[0056] Such differentiated timing settings can serve as a crucial factor in satisfying both the characteristics of query data and user expectations. In the case of message formats, real-time responses enable users to quickly obtain necessary information, while in the case of post formats, providing responses that fully reflect the context of the thread ensures the appropriateness of the conversational flow and information delivery within the post. Through this, users receive responses tailored to each data format, thereby further enhancing the reliability and usability of medical information.

[0057] As an optional embodiment, the server (120) may provide a response at an appropriate time by considering the temporal context and whether other users in the thread have posted comments, in the process of providing response data for query data in the form of a post.

[0058] Specifically, if user query data is entered in the form of a post and another user has posted a comment on the thread prior to the second point in time, the server (120) can provide response data at the third point in time by reflecting this context. Through this, the server (120) can optimize the user experience by considering the content of the existing comments and providing a response suitable for the flow of the thread.

[0059] When another user's comment is posted in a thread, it may be important to generate a customized response that reflects the content of the comment, rather than simply providing a response at a set time.

[0060] In addition, the server (120) can analyze the conversation flow of the thread and the content of existing comments to provide information suitable for the user's query and the overall context of the thread. Furthermore, by providing response data at a third point in time, the natural flow of the thread can be maintained, and more useful information can be delivered to the user by preventing the provision of unnecessary or repetitive information.

[0061] This context-based response delivery method can provide flexibility to provide appropriate responses even in environments where user interaction is active. In particular, if another user has already provided relevant information in a comment, the server (120) can improve the quality of information in the thread by generating additional information or a complementary response based on this. This enables an advanced service that goes beyond a simple chronological response delivery method, understands the context, and provides a response at an appropriate time.

[0062] As an optional embodiment, the server (120) can provide more accurate and suitable information to the user by updating the response data by reflecting comments from other users posted in the thread while providing response data in response to query data in the form of a post.

[0063] Specifically, if user query data is entered in the form of a post and another user's comment is posted in the thread before the second point in time, the server (120) can update response data based on the content of the comment to generate optimal information.

[0064] In addition, since comments from other users often provide new information or perspectives, the server (120) can improve the quality of information in the thread by analyzing them and reflecting them in the response data. For example, if a comment posted in the thread contains information directly related to the user's query or content that conflicts with existing responses, the server (120) can supplement or modify existing response data based on the comment to provide information that the user can trust. This allows for the simultaneous assurance of both the accuracy and contextual relevance of the response.

[0065] This update process can be carried out by having the AI ​​analyze the content of the comment and evaluate its relevance to the query, and then reconstruct the response data into the most suitable form. Through this, the server (120) can generate optimized information by reflecting real-time changing information of the thread, rather than simply providing existing data.

[0066] As an optional embodiment, the server (120) can simultaneously satisfy the need for user privacy and information sharing by including a function that can flexibly switch between private and public states in the process of providing the generated response data to the user.

[0067] Specifically, the generated response data can be provided in a private state by default, exposed only to the user, and can be switched to a public state to be shared with other users at the user's discretion. Through this, users can properly manage their query and response data while also providing useful information to other users by making it public as needed.

[0068] Response data provided in a private state ensures the protection of sensitive information related to users' personal inquiries and allows users to control whether their data is exposed to others. This enables users to use the medical Q&A service with peace of mind and provides trust regarding the protection of personal privacy. On the other hand, response data switched to a public state can be utilized as useful information for other users as well, enabling community-based information sharing and learning.

[0069] Users can switch response data to a public state according to their needs and intentions, thereby increasing the usability of the service. For example, if a user determines that their query addresses a common problem, they can make the corresponding response data public so that it can be used as a reference by other users. This process enables user-centric information management and improves overall service quality through information sharing within the platform.

[0070] As an optional embodiment, the server (120) may include a function to automatically switch the generated response data to a public state when a user adopts the response data during the process of switching the response data from a private state to a public state. In particular, when the user's query data is entered in the form of a post, the response data for the post is provided in a private state, and when the user judges it to be valid information and adopts it, the response data is switched to a public state so that other users can also view it. Through this, the server (120) can strengthen community-based information sharing while ensuring the suitability of the information.

[0071] The conversion of response data to public status is based on user judgment, enabling users to effectively manage their query and response data within the service. Since a user’s adoption of response data signifies trust that the information is relevant and useful to the query, converting this data to a public state allows it to be provided to other users as reliable information. This process does not merely disclose information; rather, it ensures the reliability of the information through the adoption process.

[0072] Furthermore, once response data is made public, it can be utilized as a reference by other users interested in the same topic. This goes beyond simply providing personal responses and contributes to sharing useful information within the community and enhancing the usability of the service. By sharing their questions and answers when needed, users can collaboratively create solutions for common interests.

[0073] As an optional embodiment, to enhance the reliability of the response data provided to the user, a technology may be included that provides the user with the query-answer data and medical academic data referenced in generating the response data together.

[0074] The server (120) may include a process of mapping reference data used to generate part of the response data during the process of providing response data and clearly displaying it to the user. This technology is designed so that the user can easily verify the basis of the response data provided, thereby contributing to ensuring transparency of medical information and strengthening user trust.

[0075] Specifically, when a user makes a query, the AI ​​can generate appropriate response data by searching for and analyzing information related to the user's query from pre-stored question-and-answer data and medical academic data. Some of this response data may be derived from specific academic papers or verified medical databases (130).

[0076] Additionally, the server (120) can map reference data and provide it to the user along with response data. The user interface may clearly indicate the source of the reference data along with the response data, or allow the user to check details through a hyperlink.

[0077] This mapping and delivery process transparently demonstrates to users that the response data from AI-based medical Q&A services is not merely a generated outcome, but is based on reliable sources. Through this, users can more actively utilize AI response data in their medical decision-making processes and secure the reliability and scientific basis of medical information.

[0078] As an optional embodiment, the server (120) may include a technique for clearly indicating the relationship with reference data referenced in generating the data during the process of providing response data to the user, and visually distinguishing and providing this information to the user if there is conflicting information between the referenced data.

[0079] Specifically, the server (120) displays reference data related to a portion of the response data through a first object, and if there are conflicting contents among the reference data, it can visually distinguish and provide the conflicting contents through a second object. This can increase the reliability of information for the user and support the medical decision-making process by providing various perspectives simultaneously.

[0080] Additionally, when a user queries the daily recommended amount of Vitamin D, the AI ​​generates response data by searching for information related to the question in the question-and-answer data and medical academic data. In this process, one academic data may contain conflicting information, such as one recommending 800 IU while another recommending 1000 IU. The server (120) can clearly provide this information to the user along with the response data. A portion of the response data may be designed to visually link and display the reference data that served as the basis for the recommended amount through a first object, and if there is conflicting content, to use a second object to intuitively distinguish the conflicting information.

[0081] The first and second objects can be implemented in various visual ways. For example, the first object can be displayed by linking evidence within the response data via hyperlinks or comments, while the second object can distinguish conflicting information using separate accent colors, warning icons, or pop-up notifications. Through these objects, users can easily recognize conflicting information and make medical decisions by comparing the sources and details of each piece of evidence.

[0082] This enables users to understand various perspectives on medical information and clearly identify the reasons and grounds for conflicting information. In particular, by distinguishing and providing conflicting data, it enhances the transparency of medical information for users and strengthens the reliability of AI-based medical Q&A services.

[0083] FIG. 3 is a flowchart illustrating an AI-based medical Q&A service method according to another embodiment of the present invention.

[0084] FIG. 3 describes an AI-based medical Q&A service method according to another embodiment of the present invention, in which a server (120) describes a process for efficiently processing medical data and providing optimized information to a user.

[0085] First, the server (120) can perform a process of refining and compressing data to efficiently manage and analyze original medical data composed of various formats (S310). This process can organize various forms of medical data, such as text, images, and structured data, to reduce their size and lay the foundation for rapid access to information needed in subsequent processes. Such data compression may be essential for efficiently processing large-scale medical information and increasing the speed of analysis.

[0086] Next, the server (120) can summarize the compressed data in a user-friendly form by removing unnecessary content and extracting key information based on the compressed data (S320). In this process, the focus can be on providing information that the user can immediately understand, such as symptom summaries and major treatments, from long medical papers or complex data. The summarized data ensures conciseness and accuracy, and allows the user to utilize the medical data more easily.

[0087] Next, the server (120) can analyze the summarized data and use natural language processing (NLP) technology to extract the main topics and key keywords of the information (S330). The keywords generated in this process are used as important data for information retrieval and user-customized recommendations, and can play a key role in selecting information related to the user's query data or search patterns.

[0088] Finally, the server (120) can analyze the extracted keywords and the past search history, preferences, and query context of user data to recommend highly relevant information to the user through the user terminal (110) (S340). This process is performed by the server (120) and can provide customized recommendation information by precisely filtering data that meets user requirements. Through this, the user terminal (110) can display recommendation data in a simple and intuitive manner and help the user quickly access the information they need.

[0089] FIG. 4 is an example of a screen showing the execution of an AI-based medical Q&A service according to one embodiment of the present invention.

[0090] As shown in FIG. 4(a), an exemplary configuration of a navigation button (410) that allows for effective exploration and utilization of question and answer data in a mobile user interface of an AI-based medical Q&A service can be shown. The interface can be designed so that the user can input question data, check generated response data, and conveniently explore additional detailed information.

[0091] The upper part of Fig. 4(a) displays the AI ​​response to the user-entered query "How to manage anxiety in pediatric patients?", and the response data can be organized in a user-friendly manner. In particular, the response data can present key content first and then provide related additional information in detail step by step.

[0092] The navigation buttons (410) located at the bottom can be provided as navigation tools that allow the user to quickly move to a specific area within the generated response data. Clicking the "More Than" button scrolls the user from the answer basis to the More Than area, where additional evidence or references of the response data can be viewed. Clicking the "KCI" button scrolls the user from the answer basis to the KCI area, where detailed information provided by relevant academic data or authoritative sources can be viewed. Through this button configuration, the user can efficiently navigate complex medical information and quickly access the desired information.

[0093] Additionally, the move button (410) is designed to operate efficiently even on a limited screen size in a mobile environment, allowing the user to easily select and move to a specific area through touch interaction. This allows the user to explore information step-by-step or check additional supporting materials, thereby increasing the reliability of the response data.

[0094] FIG. 5 is an example of a comment screen that executes an AI-based medical Q&A service according to one embodiment of the present invention.

[0095] As shown in FIG. 5(a), the comment area (510) of the AI-based medical Q&A service can be visually explained, showing the interaction between users and the process of the AI ​​providing responses.

[0096] The comment area (510) can be configured as an interface that allows users to check responses provided by AI, medical professionals, general users, and other users regarding a query entered by a user.

[0097] The first and second comments were written by actual users and can provide personal experiences or professional advice regarding the user's query. In particular, the display of the author's level (Lv.4) indirectly indicates the user's credibility and expertise, thereby providing a criterion for judging the quality of the comments. The written comments include specific solutions to the query and advice tailored to the patient's situation, providing practical information that users can actually refer to.

[0098] The third comment is an AI-based response labeled Moreden AI (beta), which displays answers generated by analyzing pre-stored Q&A data and medical academic data. For example, it concisely presents highly reliable information, such as implant adaptation methods based on the structure and occlusal status of the mandibular anterior teeth, and this information can assist users in making actual medical decisions.

[0099] Additionally, the comment area (510) may include a recommendation feature and a reply feature. The recommendation feature is an interface that allows users to evaluate the usefulness of comments, which can help other users quickly find useful information. The reply feature can enable in-depth discussion on medical issues by activating additional discussion and interaction among users. Furthermore, a report button is included at the top right of the comment area, allowing users to report comments containing inappropriate or unreliable information, thereby maintaining the quality of information in the comment area.

[0100] Therefore, Figure 5(a) specifically illustrates how an AI-based medical Q&A service enables information exchange between users and AI, the provision of reliable responses, and interaction among users. Through this, users can receive practical and reliable assistance regarding medical issues through the AI ​​and the user community.

[0101] FIG. 6 is an example of another figure showing an AI-based medical Q&A service implemented according to one embodiment of the present invention.

[0102] As shown in Fig. 6(a), the user interface (UI) of an AI-based medical Q&A service can be shown, and the configuration can be demonstrated that integrates the functions of providing responses to user queries and searching for related academic materials.

[0103] The interface is composed of an answer area (610), a related paper button (620), a more than area (630), and a KCI (Korea Citation Index) area (640), and can be designed to provide the user with clear and in-depth information regarding the query.

[0104] The answer area (610) may be a main space for displaying response data generated by the AI ​​in response to a query entered by the user. In this area, the AI ​​can systematically organize the cause of the problem, solution methods, prevention strategies, etc., in response to the query and provide them to the user.

[0105] Furthermore, this area is designed with a user-friendly UI to help users intuitively check and utilize information. The response data contains concise yet practical information, enabling direct assistance in solving users' problems.

[0106] The related paper button (620) at the bottom of the answer area can provide a function that helps the user explore related academic materials along with the AI ​​response data. When the button is clicked, papers or research materials related to the query are displayed to complement the response data and enhance the reliability of the information. Through this, the user can expand the depth of information by verifying additional academic evidence, rather than simply stopping at the AI ​​response.

[0107] The Moreden area (630) is a space where the user can explore additional information related to the query, and this area provides the user with additional context regarding the query and can help expand and deepen their understanding of medical knowledge.

[0108] The KCI (Korea Citation Index) area (640) can provide papers and research materials related to user queries based on domestic academic data. This area displays the latest papers, and users can verify the academic basis of the AI ​​response data through this. Through this, the KCI area (640) can help users obtain objective and verified information based on reliability and expertise.

[0109] FIG. 7 is a diagram showing a computer system according to one embodiment of the present invention.

[0110] Referring to FIG. 7, a computer system (1000) may include one or more processors (1010), memory (1030), user interface input device (1040), user interface output device (1050), and storage (1060) that communicate with each other via a bus (1020). Additionally, the computer system (1000) may further include a network interface (1070) connected to a network (1080). The processor (1010) may be a central processing unit or a semiconductor device that executes processing instructions stored in memory (1030) or storage (1060). Memory (1030) and storage (1060) may be various forms of volatile or non-volatile storage media. For example, memory may include ROM (1031) or RAM (1032). Explanation of the symbols

[0111] 110: User terminal 120: Server 130: Database 1000: Computer System 1010: Processor 1020: Bus 1030: Memory 1031: Rome 1032: RAM 1040: User interface input device 1050: User interface output device 1060: Storage 1070: Network Interface 1080: Network

Claims

Claim 1 A method for an AI-based medical Q&A service comprising: receiving user input data including user query data from a user terminal; generating AI-based response data using at least one of previously stored question-response data and medical academic data as a response to the user input data; and providing the response data to the user through the user terminal. Claim 2 In claim 1, the query data is in the form of a message entered through a chat window or a post posted as the start post of a thread, and the step of providing the response data includes the step of providing the response data as a response message or a comment of the thread corresponding to the format of the query data, an AI-based medical Q&A service method. Claim 3 In paragraph 2, the step of providing the response data comprises: providing the response message at a first time point when the query data is in the message format; or providing the comment at a second time point later than the first time point when the query data is in the post format. Claim 4 In paragraph 3, the step of providing the response data includes, if the query data is in the form of a post and a comment by another user is posted in the thread before the second time point, the step of providing the comment at a third time point later than the second time point. Claim 5 In claim 4, the step of generating the response data includes the step of updating the response data based on the comment when the query data is in the form of the post and a comment from another user is posted in the thread before the second point in time. Claim 6 The AI-based medical Q&A service method according to claim 1, wherein the step of providing the response data comprises: providing the response data in a private state so as to be exposed only to the user; and switching the response data to a public state so as to be exposed to other users according to the user's selection. Claim 7 In claim 6, the step of switching to a public state includes the step of switching the response data from a private state to a public state when the query data is in the form of a post and the user adopts the response data. Claim 8 The AI-based medical Q&A service method according to claim 1, wherein the step of providing the response data comprises: a step of mapping at least one of the question-answer data and medical academic data used to generate a part of the response data as reference data; and a step of providing the reference data together with the response data. Claim 9 In claim 8, the step of providing the response data comprises: a step of displaying a first object so that the reference data is associated with a part of the response data to which the reference data is mapped; and a step of displaying a second object distinguished from the first object so that, if there is conflicting content between the mapped reference data, the conflicting reference data is associated with a part of the response data to which the response data is mapped, in the AI-based medical Q&A service method. Claim 10 An AI-based medical Q&A service device comprising: a memory on which at least one program is recorded; and a processor for executing said program, wherein the program comprises instructions for performing the steps of: receiving user input data including user query data from a user terminal; generating AI-based response data using at least one of previously stored question-response data and medical academic data as a response to said user input data; and providing said response data to the user through said user terminal.

Images