Method and system for providing exercise program by using large language model

Abstract

The present invention relates to a method for providing an exercise program by using a large language model, the method comprising the steps of: providing, to a user terminal, at least one questionnaire related to user's indications; receiving, from the user terminal, response data to the questionnaire; assessing, by using a pre-trained motion assessment model, exercise performance ability of the user from exercise motion data received from the user terminal; by using at least one of the response data to the questionnaire or the exercise performance ability assessment result, generating a prompt to be input to a large language model; generating an exercise program related to the user's indications through the large language model by inputting the prompt to the large language model; and identifying the exercise program and providing the identified exercise program to the user terminal.

Description

Method and System for Providing Exercise Programs Using Large Language Models

[0001] The present invention relates to a method and system for providing an exercise program using a large language model.

[0002] Recently, with the rapid advancement of artificial intelligence (AI) technology, generative models capable of natural conversation with humans (e.g., ChatGPT) have emerged. In particular, generative models that include a Large Language Model are demonstrating innovation in the AI ​​market by showcasing technological capabilities that allow them to communicate naturally, almost like humans, and provide fast and accurate information, unlike traditional chatbots that are manually built and provide only limited answers.

[0003] Furthermore, the utilization of artificial intelligence (AI) technology in the medical industry is rapidly expanding. In particular, there is growing interest in AI technology that uses large language models to provide personalized, non-face-to-face care to patients requiring consistent management and rehabilitation during treatment.

[0004] For example, large language models can be effectively utilized for the management and rehabilitation of musculoskeletal disorders. Musculoskeletal disorders refer to pain or injury occurring in the musculoskeletal system, including muscles, nerves, tendons, ligaments, bones, and surrounding tissues. As a principle, the treatment of musculoskeletal disorders should begin with less invasive procedures; non-pharmacological conservative treatments (e.g., exercise therapy and education, cognitive therapy, or relaxation therapy) should be implemented first, followed by pharmacological treatment and surgical treatment in sequence. Treatment guidelines strongly recommend non-pharmacological conservative treatment for musculoskeletal disorders, and active research on methods for implementing such treatments is being conducted, primarily in the United States and Europe. However, since continuous treatment and rehabilitation are crucial for non-pharmacological conservative treatment, the need for patients to visit the hospital frequently poses a significant burden.

[0005] To address these issues and facilitate the consistent management and rehabilitation of musculoskeletal disorders, there is a need to provide user-customized exercise programs remotely using large language models.

[0006] The present invention is intended to provide a method and system capable of providing a customized exercise program to a user using a large language model.

[0007] More specifically, the present invention aims to provide a method and system for providing an exercise program using a large language model capable of generating a user-customized exercise program based on response data to a survey provided to a user terminal and an evaluation of the user's exercise performance ability.

[0008] Furthermore, the present invention aims to provide a method and system for providing an exercise program using a large language model, which can update an exercise program based on feedback received from a medical staff terminal and provide an exercise program approved by the medical staff terminal.

[0009] To solve the problem described above, the present invention proposes a method for generating a user-customized exercise program by interacting with a user and medical staff remotely using a large language model. The method for providing an exercise program using a large language model according to the present invention may include the steps of: providing at least one survey related to the user's indication to a user terminal; receiving response data for the survey from the user terminal; evaluating the user's exercise performance ability from the exercise motion data received from the user terminal using a pre-trained motion evaluation model; generating a prompt to be input into a large language model using at least one of the response data for the survey and the exercise performance evaluation result; inputting the prompt into the large language model to generate an exercise program related to the user's indication through the large language model; transmitting a request for review of the exercise program generated through the large language model to a medical staff terminal; receiving feedback regarding the exercise program from the medical staff terminal in response to the review request; specifying the exercise program based on the feedback; and providing the specified exercise program to the user terminal.

[0010] Furthermore, the above exercise program is configured to include a plurality of exercise modules, each of the plurality of exercise modules is matched to a different type of exercise and may include at least one exercise motion content related to the type of exercise matched to each exercise module.

[0011] Furthermore, the step of receiving the feedback may include providing a service page associated with the review request to the medical staff terminal, displaying information associated with the exercise program on the service page, and receiving feedback from the medical staff regarding the exercise program through the service page.

[0012] Furthermore, the information associated with the exercise program includes an exercise module list for the plurality of exercise modules constituting the exercise program, and the exercise module list may include at least one of the exercise motion content included in each of the plurality of exercise modules and a review icon indicating whether the medical staff has reviewed each of the plurality of exercise modules.

[0013] Furthermore, the above feedback may be at least one of a first feedback approving the exercise program and a second feedback requesting modifications to the exercise program.

[0014] Furthermore, in the step of specifying the exercise program, different data processing processes for the exercise program are performed according to the feedback received from the medical staff terminal, and the different data processing processes may include at least one of a first process of specifying the generated exercise program and providing the specified exercise program to the user terminal based on receiving the first feedback from the medical staff terminal, and a second process of modifying the exercise program based on receiving the second feedback from the medical staff terminal.

[0015] Furthermore, the second process may include the steps of: receiving a medical staff query related to at least one specific exercise module among the plurality of exercise modules constituting the exercise program from the medical staff terminal via the service page; processing information regarding the specific exercise module in which the medical staff query was received and the medical staff query as input to the large language model to obtain an answer from the large language model related to the specific exercise module; and modifying the specific exercise module using the answer from the large language model related to the specific exercise module.

[0016] Furthermore, the step of modifying the specific exercise module comprises: providing at least one candidate exercise motion content associated with the specific exercise module to be modified to the medical staff terminal based on the medical staff query related to the specific exercise module; identifying, from the medical staff terminal, the specific exercise motion content included in the specific exercise module and the specific candidate exercise motion content to be changed among the at least one candidate exercise motion content; and changing the specific exercise motion content included in the specific exercise module to the specific candidate exercise motion content. This characterizes a method for providing an exercise program using a large language model.

[0017] Furthermore, the above different types of exercise may be at least one of a warm-up exercise type, a main exercise type, and a cool-down exercise type.

[0018] Furthermore, the above-mentioned at least one survey may include at least one of a multiple-choice survey consisting of a question related to the user's indication and a plurality of selection items corresponding to each of a plurality of different responses to the question, and a conversational survey capable of receiving natural language input from the user terminal in relation to the user's indication.

[0019] Furthermore, the step of receiving response data for the above survey may include the step of receiving the selection-type survey response data from the user terminal, which includes a response matched to an item selected by user input among the plurality of selection items, and the step of receiving the conversational survey response data from the large language model by processing the natural language input entered in the conversational survey as input to the large language model.

[0020] Furthermore, the step of evaluating the exercise performance ability of the user may include receiving exercise motion data regarding a specific posture of the user captured by a camera equipped in the user terminal, and analyzing at least one of the user's static posture, joint range of motion, balance ability, and core strength using the previously learned motion evaluation model.

[0021] Furthermore, the step of generating a prompt to be input into the large language model may include the step of specifying a body part in which the user's joint range of motion satisfies a preset condition, and the step of generating the prompt so that exercise motion content related to the specified body part is included in the exercise program.

[0022] Meanwhile, the exercise program providing system using a large language model according to the present invention includes a control unit that provides at least one survey related to the user's indication to a user terminal, and a communication unit that receives response data for the survey from the user terminal. The control unit evaluates the user's exercise performance ability from exercise motion data received from the user terminal using a pre-trained motion evaluation model, generates a prompt to be input into a large language model using at least one of the response data for the survey and the exercise performance ability evaluation result, inputs the prompt into the large language model to generate an exercise program related to the user's indication, transmits a request for review of the exercise program generated through the large language model to a medical staff terminal, receives feedback regarding the exercise program from the medical staff terminal in response to the review request, specifies the exercise program based on the feedback, and can provide the specified exercise program to the user terminal.

[0023] Meanwhile, the program is executed by one or more processes in an electronic device and is stored on a computer-readable recording medium, and the program may include instructions for performing the steps of: providing at least one survey related to the user’s indication to a user terminal; receiving response data for the survey from the user terminal; evaluating the user’s exercise performance ability from the exercise motion data received from the user terminal using a pre-trained motion evaluation model; generating a prompt to be input into a large language model using at least one of the response data for the survey and the exercise performance evaluation result; inputting the prompt into the large language model to generate an exercise program related to the user’s indication through the large language model; transmitting a request for review of the exercise program generated through the large language model to a medical staff terminal; receiving feedback regarding the exercise program from the medical staff terminal in response to the review request; specifying the exercise program based on the feedback; and providing the specified exercise program to the user terminal.

[0024] The method and system for providing an exercise program using a large language model according to the present invention can generate a user-customized exercise program by utilizing response data to a survey provided to a user terminal and the results of an evaluation of the user's exercise performance ability. Through this, the present invention can generate an exercise program of a difficulty level that considers the user's current condition, thereby maximizing the rehabilitation effect for the user's indications.

[0025] Furthermore, the method and system for providing an exercise program using a large language model according to the present invention can provide an exercise program with guaranteed expertise and safety by utilizing a large language model and reflecting feedback from medical staff regarding the exercise program.

[0026] Furthermore, the method and system for providing an exercise program using a large language model according to the present invention can provide a generated exercise program to a user terminal remotely based on feedback from medical staff. Through this, patients can receive rehabilitation treatment without having to visit a hospital located far away, thereby enabling easy access to rehabilitation treatment. Additionally, medical staff can conveniently monitor the patient's rehabilitation exercises through electronic devices and provide monitoring-based feedback to enhance the effectiveness of the patient's exercise therapy.

[0027] FIG. 1 is a conceptual diagram illustrating an exercise program providing system using a large language model according to the present invention.

[0028] FIG. 2 is a flowchart illustrating a method for providing an exercise program using a large language model according to the present invention.

[0029] FIGS. 3a and FIGS. 3b are conceptual diagrams for explaining the survey process according to the present invention.

[0030] FIGS. 4a and FIGS. 4b are conceptual diagrams for explaining the exercise performance evaluation process according to the present invention.

[0031] FIG. 5 is a conceptual diagram illustrating the prompt generation process according to the present invention.

[0032] FIG. 6 is a conceptual diagram illustrating the process of generating an exercise program using a large language model according to the present invention.

[0033] FIGS. 7, FIGS. 8a to 8c, FIGS. 9a, and FIGS. 9b are conceptual diagrams for explaining the process of generating a final exercise program using a large language model in a medical staff terminal according to the present invention and providing it to a user terminal.

[0034] Hereinafter, embodiments disclosed in this specification will be described in detail with reference to the attached drawings. Identical or similar components are assigned the same reference number regardless of the drawing symbols, and redundant descriptions thereof will be omitted. The suffixes "module" and "part" used for components in the following description are assigned or used interchangeably solely for the ease of drafting the specification and do not have distinct meanings or roles in themselves. Furthermore, in describing the embodiments disclosed in this specification, if it is determined that a detailed description of related prior art could obscure the essence of the embodiments disclosed in this specification, such detailed description will be omitted. Additionally, the attached drawings are intended only to facilitate understanding of the embodiments disclosed in this specification; the technical concept disclosed in this specification is not limited by the attached drawings, and it should be understood that they include all modifications, equivalents, and substitutions that fall within the spirit and technical scope of the present invention.

[0035] Terms including ordinal numbers, such as first, second, etc., may be used to describe various components, but said components are not limited by said terms. These terms are used solely for the purpose of distinguishing one component from another.

[0036] When it is stated that one component is "connected" or "connected" to another component, it should be understood that while it may be directly connected or connected to that other component, there may also be other components in between. On the other hand, when it is stated that one component is "directly connected" or "directly connected" to another component, it should be understood that there are no other components in between.

[0037] A singular expression includes a plural expression unless the context clearly indicates otherwise.

[0038] In this application, terms such as “comprising” or “having” are intended to specify the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0039] The present invention relates to a method and system for providing an exercise program using a large language model. More specifically, the present invention relates to a method and system for providing a user-customized exercise program that incorporates feedback from medical professionals regarding the exercise program, using a large language model. Here, "large language model (LLM)" may refer to an artificial intelligence model capable of understanding and generating natural language by learning from a vast amount of data.

[0040] The “exercise program” according to the present invention may refer to an exercise plan provided to a user terminal in relation to the user’s indication. Here, “indication” refers to a symptom or clinical situation requiring specific treatment or examination, and may be understood as the user’s disease or symptom. For example, in the present invention, if the user has an indication related to a specific body part (e.g., lower back (lumbar spine)), an exercise program for the rehabilitation of the indication possessed by the user may be generated.

[0041] More specifically, an exercise program according to the present invention may be configured to include at least one exercise module. Here, an “exercise module” may refer to a constituent unit of an exercise program that includes at least one exercise motion content. In this case, the “exercise motion content” may include various information such as a name for a specific exercise motion, exercise time information, exercise difficulty information, exercise type information, exercise body part information, and exercise posture information. Meanwhile, the “exercise motion” described in the present invention refers to a gesture (movement) performed during the process of performing exercise, and may be used interchangeably with terms such as “movement,” “action,” “movement,” and “gesture.”

[0042] For convenience of explanation, the present invention focuses on “indications related to musculoskeletal disorders,” but is not necessarily limited thereto. As an example, the exercise program described in the present invention may be for the treatment of users requiring rehabilitation treatment for musculoskeletal disorders and users suffering from various diseases (e.g., cancer, diabetes, hypertension, etc.).

[0043] Furthermore, the present invention may provide an exercise program necessary for health promotion in daily life, rather than rehabilitation exercises for therapeutic purposes related to the user's indications. For example, the exercise according to the present invention is not limited to any specific purpose and may be an exercise performed for various purposes, such as rehabilitation exercises, fitness exercises, ball sports, or dance exercises, for therapeutic purposes, health promotion purposes, or beauty purposes.

[0044] Furthermore, the exercise program according to the present invention may refer to a set of exercise modules for various exercises, such as rehabilitation exercises, fitness exercises, ball sports, and dance exercises, for various purposes including therapeutic, health promotion, and beauty purposes, and may be understood not to be limited to a specific category of exercise. That is, there is no limitation on the type of exercise according to the present invention, nor is there a limitation on the location of the exercise, such as indoor or outdoor exercise.

[0045] In the foregoing, the provision of an exercise program using a large language model according to the present invention has been generally described, and this can be implemented by an exercise program provision system using a large language model described below. Below, with reference to FIG. 1, an exercise program provision system using a large language model according to the present invention will be described in detail. FIG. 1 is a conceptual diagram for explaining an exercise program provision system using a large language model according to the present invention.

[0046] As illustrated in FIG. 1, the generative model-based exercise program providing system according to the present invention (hereinafter referred to as the “exercise program providing system,” 100) may include at least one of a communication unit (110), a storage unit (120), and a control unit (130). At this time, the exercise program providing system (100) according to the present invention is not limited to the components described above and may further include components that perform the same or similar roles as the functions described in the specification. Meanwhile, the exercise program providing system (100) according to the present invention may be implemented as an application or software. The exercise program providing system (100) implemented as software in this manner may be downloaded via a program (e.g., Play Store) that allows the application to be downloaded on a user terminal (10) and a medical staff terminal (20), or implemented via an initial installation program on the user terminal (10) and the medical staff terminal (20). In this case, the communication unit (110), storage unit (120), and control unit (130) according to the present invention can be utilized as components of a user terminal (10) and a medical staff terminal (20). In the present invention, the user terminal (10) can be understood to mean an application installed on the user terminal (10). Such an application (or software) can be understood as a component of the exercise program providing system (100) according to the present invention.

[0047] In the present invention, the user terminal (10) and the medical staff terminal (20) may also be referred to as a "mobile terminal" or "electronic device," and the user terminal (10) described in this specification may include a mobile phone, a smartphone, a smart TV, a laptop computer, a digital broadcasting terminal, a PDA (personal digital assistants), a PMP (portable multimedia player), a navigation device, a slate PC, a tablet PC, an ultrabook, a wearable device (e.g., a smartwatch, a smart glass, a head-mounted display), etc.

[0048] More specifically, the user terminal (10) and the medical staff terminal (20) according to the present invention are not limited to electronic devices with an application activated, but may refer to electronic devices connected to electronic devices with an application activated. As an example, based on the fact that the user terminal (10) according to the present invention is a smartphone, the user terminal (10) may refer to a smart TV connected to said smartphone.

[0049] Meanwhile, the exercise program providing system (100) may exist inside a server (hereinafter referred to as the server) built to perform a specific purpose (e.g., providing an exercise program), or it may exist as a separate device from the server. When the exercise program providing system (100) exists inside the server, the exercise program providing system (100) according to the present invention may provide a user-customized exercise program through at least one component among a communication unit (110), a storage unit (120), and a control unit (130) located inside the server, or through a module that performs a function similar to each of the above components. In this case, the application may provide the exercise program to a user terminal (10) on which the application is installed through communication with the server. Furthermore, the exercise program providing system (100) according to the present invention may provide the exercise program according to the present invention to a user terminal (10) by linking with a plurality of different external servers.

[0050] A user (or patient, U) and a medical staff member (or doctor, D) according to the present invention may possess an account registered in the exercise program providing system (100) according to the present invention. For convenience of explanation, the account of the user (or patient) in this specification is referred to as the "user account (or patient account)." The "account" described above may be created through a page linked to the exercise program providing system (100). Alternatively, the "account" may be created on at least one other server (e.g., a medical staff server) linked to the exercise program providing system (100) according to the present invention. Accordingly, in this specification, without distinguishing the server where the account was issued, all accounts based on the exercise program providing system (100) according to the present invention are referred to as "accounts already registered in the exercise program providing system (100) according to the present invention."

[0051] Furthermore, a user (U, or patient) according to the present invention may receive an exercise program related to the indications of the user (U) through an application or webpage provided by the exercise program providing system (100) according to the present invention.

[0052] Meanwhile, the doctor can provide feedback on the exercise program to be provided to the user (U) through the medical staff terminal (20). At this time, the doctor (D) may possess a doctor account already registered in the exercise program providing system (100) according to the present invention. In this specification, the user terminal logged in with the doctor account is referred to as the medical staff terminal (20). As an example, the exercise program providing system (100) according to the present invention may receive an exercise program approved by the doctor (D) for the user (U) by linking with a medical staff server.

[0053] Furthermore, the “medical staff” described in the present invention refers to persons employed at a medical institution (e.g., a hospital) and may include, for example, at least one of a doctor, a nurse, or a physical therapist. For convenience of explanation, the present invention describes medical staff by citing doctors and physical therapists as examples. However, medical staff are not limited thereto, and any user employed at a medical institution to provide a user-customized exercise program may be considered medical staff according to the present invention.

[0054] An exercise program providing system (100) according to the present invention can collect sensing information including at least one of voice data and video data based on a microphone, camera, and sensor unit provided in a user terminal (10). Here, the sensor unit may include at least one sensor among an infrared sensor, a LiDAR sensor, an accelerometer, an illuminance sensor, a proximity sensor, a position sensor, a face recognition sensor, an iris scanner, a heart rate sensor, a touch sensor, and a pressure sensor.

[0055] According to the present invention, the communication unit (110) may be connected via a wireless or wired network to a user terminal (10), a medical staff terminal (20), an LLM server (140), a central server, a device, and at least one network, and may be configured to receive or transmit overall data and information necessary for the operation of the posture estimation system (100) according to the present invention.

[0056] Specifically, the communication unit (110) can transmit at least one survey to the user terminal (10). Furthermore, the communication unit (110) can receive response data for at least one survey provided to the user terminal (10). Here, “receiving response data” may mean receiving an input signal (or selection signal) corresponding to a user input input through the user terminal (10).

[0057] The communication unit (110) can receive an exercise program already stored in a database (or DB, 200). Furthermore, the communication unit (110) can receive patient history data from the database (200).

[0058] Additionally, the communication unit (110) can receive user exercise motion data captured from the user terminal (10). Here, the exercise motion data may refer to at least one of an image and a video of the user's exercise motion captured by a camera equipped in the user terminal (10).

[0059] Furthermore, the communication unit (110) can receive feedback from the medical staff terminal (20) including at least one of approval and modification of the exercise program.

[0060] The communication unit (110) may include at least one communication module capable of wireless communication and wired communication between the posture estimation system (100) and the communication target. Additionally, the communication unit (110) may include a communication module that connects the exercise program providing system (100) to at least one network.

[0061] Meanwhile, the communication unit (110) can support various communication methods depending on the communication standard of the communicating device. For example, the communication unit (110) may be configured to perform communication using at least one of the following technologies: WLAN (Wireless LAN), Wi-Fi (Wireless-Fidelity), Wi-Fi (Wireless Fidelity) Direct, DLNA (Digital Living Network Alliance), WiBro (Wireless Broadband), WiMAX (World Interoperability for Microwave Access), HSDPA (High Speed ​​Downlink Packet Access), HSUPA (High Speed ​​Uplink Packet Access), LTE (Long Term Evolution), LTE-A (Long Term Evolution-Advanced), 5G (5th Generation Mobile Telecommunication), Bluetooth (Bluetooth™ Frequency Identification), Infrared Communication (Infrared Data Association; IrDA), UWB (Ultra-Wideband), ZigBee, NFC (Near Field Communication), Wi-Fi Direct, and Wireless USB (Wireless Universal Serial Bus).

[0062] Next, the storage unit (120) may be configured to store various information related to the present invention. In the present invention, the storage unit (120) may be provided in the exercise program providing system (100) itself, or alternatively, at least a part of the storage unit (120) may mean a database (Database: DB, 200).

[0063] The storage unit (120) may include one or more non-transient computer-readable storage media that can be read and / or accessed by at least one processor. One or more computer-readable storage media may include volatile and / or non-volatile storage components such as optical, magnetic, organic, or other memory or disk storage devices. In some examples, the storage unit (120) may be implemented using a single physical device (e.g., one optical, magnetic, organic, or other memory or disk storage device), whereas in other examples, the storage unit (120) may be implemented using multiple physical devices. The storage unit (120) may include computer-readable instructions and additional data.

[0064] The storage unit (120) may include a storage necessary to perform at least some of the methods and techniques described in this specification and / or at least some of the functions of the device and network. That is, the storage unit (120) is sufficient as a space in which information necessary to provide an exercise program according to the present invention is stored, and the storage unit (120) can be understood as not being restricted by physical space.

[0065] Data and commands necessary for the operation of the exercise program providing system (100) according to the present invention may be stored in the storage unit (120). More specifically, data and commands necessary for the operation of the motion evaluation model (132) may be stored in the storage unit (120). Here, the motion evaluation model (132) may refer to an artificial intelligence model that analyzes the user's exercise motion performed according to the exercise performance evaluation provided to the user terminal and performs an evaluation of the exercise motion.

[0066] For example, the motion evaluation model (132) of the present invention is a posture estimation model learned using a learning data set containing position information for joint points, and can estimate the exercise posture of a user (U) from an exercise video to be analyzed. Here, “joint point” may refer to a plurality of joints of the user (U) (or a part of the user (U)’s body including joints). And, “key point” may refer to an area corresponding to each of the plurality of joint points of the user (U) in the exercise motion video.

[0067] User-related information may be stored in the storage unit (120). User-related information may include various information related to the provision of an exercise program, such as i) user ID, ii) name, iii) date of birth, iv) indications of the user, v) exercise history, vi) medical history, vii) treatment plan, etc., according to at least one of user account information and prescription information. Furthermore, user authentication information included in the user account information may be stored in the storage unit (120). Here, “user authentication information” may refer to information used in a user authentication process performed to log in to a user account on a user terminal. As an example, user authentication information may be various, such as i) ID, ii) password, iii) password pattern, iv) user fingerprint authentication information, v) facial authentication information, vi) voice authentication information, vii) iris authentication information, viii) vein authentication information, etc., set by the user.

[0068] Additionally, at least one of an exercise video and an exercise description corresponding to each exercise motion may be stored in the storage unit (120).

[0069] Meanwhile, the database (DB, 200) may be configured to store various information related to providing a user-customized exercise program. Specifically, the database (200) may store exercise movement content (e.g., exercise name, number of exercises, timing of exercises, difficulty of exercises, etc.) corresponding to each of multiple pain sites related to the indication (e.g., shoulder, elbow, wrist & hand, hip & pelvis, knee, ankle & foot, neck, back, waist, abdomen). Furthermore, the database (200) may store at least one of an exercise video and an exercise description corresponding to each exercise movement content.

[0070] Additionally, the database (200) may store patient history information for each of multiple different users. For example, the database (200) may store patient history information for each of multiple users, including at least one of i) name, ii) date of birth, iii) type of indication of the user, iv) exercise history, v) past medical history, vi) prescription information, vii) response data history information for a questionnaire, and viii) exercise prognosis information.

[0071] In the present invention, the storage unit (120) is described as existing separately from the database (200), but is not limited thereto, and the storage unit (120) may include the database (200).

[0072] Next, the control unit (130) may be configured to control the overall operation of the exercise program providing system (100) related to the present invention. The control unit (130) may include at least one of a prompt generation unit (131) and an operation evaluation model (132). Furthermore, the control unit (130) may process signals, data, information, etc. that are input or output through the components described above, or provide or process appropriate information and functions to the user.

[0073] The control unit (130) can control the output of a service page for providing an exercise program through a display unit (or touchscreen) provided in the user terminal (10) and the medical staff terminal (20). Such a service page may be output on the user terminal (10) and the medical staff terminal (20) through an application or web page installed on the user terminal (10) and the medical staff terminal (20). The service page is a page linked to the exercise program providing system (100) according to the present invention and is configured to be controlled by the exercise program providing system (100) according to the present invention.

[0074] Furthermore, if the service page is provided in the form of an application, the service page may be controlled by the CPU (Central processing unit) of the user terminal (10) on which the application is installed. In this case, the CPU of the user terminal (10) may provide a customized exercise program to the user based on information provided by the exercise program providing system (100) according to the present invention.

[0075] Meanwhile, the control unit (130) can provide at least one survey on the service page provided to the user terminal (10).

[0076] Furthermore, the control unit (130) can evaluate the user's exercise performance ability using a previously learned motion evaluation model (132). More specifically, the control unit (130) can generate an exercise performance ability evaluation result by inputting the user's exercise motion data, captured (or sensed) using at least one of the camera and sensor unit provided in the user terminal (10), into the motion evaluation model.

[0077] The control unit (130) can generate a prompt to be input to the LLM server using at least one of the response data for at least one survey provided to the user terminal (10) and at least one of the exercise performance evaluation results. At this time, the LLM server (140) according to the present invention may include at least one large language model. For example, the LLM server (140) may include at least one large language model based on at least one of GPT (Generative Pre-trained Transformer), T5 (Text-to-Text Transfer Transformer), Vision-Language Models (e.g., CLIP, Flamingo), Multimodal Generative Models (GPT-4 Multimodal, PaLM-E), BERT (Bidirectional Encoder Representations from Transformers), and LaMDA (Language Model for Dialogue Applications).

[0078] In the present invention, the LLM server (140) is described as existing separately from the attitude estimation system (100), but is not limited thereto, and the attitude estimation system (100) may be configured to include the LLM server (140). That is, the attitude estimation system (100) and the LLM server (140) according to the present invention may exist separately, or the LLM server (140) may be included in the attitude estimation system (100). For convenience of explanation, the LLM server (140) and the large language model are used interchangeably below, and the use of the large language model by the control unit (130) can be understood as using at least one large language model included in the LLM server (140).

[0079] Additionally, although the present invention describes the control unit (130) as including a prompt generation unit (131), it is not limited thereto, and the control unit (130) and the prompt generation unit (131) may exist separately. In this case, the control unit (130) can generate a prompt in conjunction with the prompt generation unit (131).

[0080] The control unit (130) according to the present invention can generate a user-customized exercise program corresponding to a prompt by using a large language model (140). Furthermore, the control unit (130) can review the exercise program based on feedback received from the medical staff terminal (20) in conjunction with the large language model (140). Here, “feedback” may refer to user input related to at least one of modification (or update) and approval of the exercise program by the medical staff. For example, it may refer to user input that modifies the exercise program using the response of the large language model to a medical staff query entered into the medical staff terminal (20).

[0081] Furthermore, the control unit (130) can provide a specific exercise program (or a final exercise program) to the user terminal (10) in response to the specific exercise program specified by the medical staff terminal (20).

[0082] Meanwhile, the exercise program providing system (100) may include one or more processors, and such processors may include one or more general-purpose processors and / or one or more special-purpose processors (e.g., digital signal processors, tensor processing units (TPUs), graphics processing units (GPUs), neural network processing units (NPUs), application integrated circuits, application semiconductors (ASICs), etc.). One or more processors may be configured to execute instructions, computer-readable instructions, and / or other instructions described herein that are stored (or included) in the storage unit (120). The exercise program providing system (100) may perform data processing described below in cooperation with memory and at least one processor. The processor may perform a series of operations and data processing using data and information stored in memory. Here, “memory” may be a component of the storage unit (120), and “processor” may be used interchangeably with the control unit (130).

[0083] In the above description, the exercise program providing system (100) of the present invention has been described, and it can be implemented based on the method of providing an exercise program using a large language model described below.

[0084] Hereinafter, with reference to FIG. 2 together with FIG. 3a, FIG. 3b, FIG. 4a, FIG. 4b, FIG. 5, FIG. 6, FIG. 7, FIG. 8a to FIG. 8c and FIG. 9, a method for providing an exercise program using a large language model according to the present invention will be described in more detail. FIG. 2 is a flowchart for explaining a method for providing an exercise program using a large language model according to the present invention, and FIG. 3a and FIG. 3b are conceptual diagrams for explaining a survey process according to the present invention. FIG. 4a and FIG. 4b are conceptual diagrams for explaining an exercise performance evaluation process according to the present invention, and FIG. 5 is a conceptual diagram for explaining a prompt generation process according to the present invention. FIG. 6 is a conceptual diagram for explaining a process of generating an exercise program using a large language model according to the present invention, and FIG. 7, FIG. 8a to FIG. 8c, FIG. 9a and FIG. 9b are conceptual diagrams for explaining a process of generating a final exercise program using a large language model at a medical staff terminal and providing it to a user terminal according to the present invention.

[0085] In the present invention, a process of providing at least one questionnaire related to the user's indication to a user terminal may be carried out (S210, see FIG. 2).

[0086] As illustrated in FIG. 3(a), the control unit (130) may provide a preliminary questionnaire (300) to the service page (1000) provided to the user terminal (10) to input a painful body part (or musculoskeletal part) in relation to the user's indication. At this time, the preliminary questionnaire (300) may include a selection type survey to select a painful body part in relation to the user's indication. Specifically, the control unit (130) may provide the user terminal (10) with a body icon corresponding to each body part in a body diagram.

[0087] Furthermore, the control unit (130) may provide guidance information (e.g., “Please select the area where you have pain”) to induce the user to select at least one body part. As an example, the control unit (130) may display body icons corresponding to each of a plurality of body parts (e.g., “neck (cervical spine), “shoulder,” “back (thoracic spine),” “waist (lumbar spine),” “elbow,” “wrist,” “knee,” “ankle,” etc.) on a body diagram provided to the user terminal (10). The user (U) may identify the body part where the pain is present through a selection survey that allows the user to select the area where the pain is present. Specifically, the control unit (130) may identify the body part corresponding to the specific icon (e.g., waist (lumbar spine)) as the user's area of ​​pain based on the user input (312) for a specific icon (311) among the plurality of body icons provided to the user terminal (10) by the user (U).

[0088] As illustrated in FIG. 3(b), the control unit (130) may provide a preliminary questionnaire (300) that allows the user to input the degree of pain on a service page (1000) provided to the user terminal (10) when the pain area is specified. At this time, the preliminary questionnaire (300) may include a selection type survey that allows the user to input the degree of pain regarding the specified pain area.

[0089] Specifically, the control unit (130) may provide a selection item (320) that allows the user (U) to input the degree of pain regarding a specific pain area on the user terminal (10). Here, “degree of pain” may refer to a Numerical Rating Scale (NRS) and may consist of numbers from 0 to 10. Specifically, the control unit (130) may provide a selection survey that includes a selection item (320) to express the degree of pain felt by the user (U) as a number.

[0090] Furthermore, the control unit (130) may provide at least one selection survey (or selection survey item) configured to allow the user (U) to directly select the user's (U) health condition in response to the user (U) entering the pain level (322) into a selection item (320) where the user (U) can input the pain level. Here, the “selection survey” may be composed of a question related to the user’s indication and a plurality of selection items corresponding to each of a plurality of different responses to the question.

[0091] The storage unit (120) stores multiple optional questionnaires for each of the multiple pain areas to check the user's symptoms for that area. These optional questionnaires can be configured differently depending on the specific pain area of ​​the user. The control unit (130) can refer to the storage unit (120) and provide the optional questionnaire for a specific pain area among the multiple body parts to the user terminal (10) to determine the user's indication status for the specific pain area.

[0092] As illustrated in FIG. 3b, the control unit (130) can conduct a multi-faceted survey on the user's indications by providing the user terminal (10) with a plurality of multiple-choice surveys (331, 332, 333, 334) and selection items corresponding to each of the multiple different responses to each of the questions related to the user's indications included in the multiple-choice surveys. For example, the control unit (130) can conduct a survey on exercise equipment that can be used during an exercise program by using a multiple-choice survey (334) that includes questions related to exercise equipment.

[0093] Furthermore, the control unit (130) can conduct an AI-based conversational survey. Here, the “conversational survey” may refer to a survey that can receive natural language input from the user terminal in relation to the user’s indication. At this time, the conversational survey may be conducted using at least one of a large language model (140) and a chatbot. Here, the at least one large language model (140) may refer to a large language model (140) based on at least one of GPT (Generative Pre-trained Transformer), T5 (Text-to-Text Transfer Transformer), Vision-Language Models (e.g., CLIP, Flamingo), Multimodal Generative Models (GPT-4 Multimodal, PaLM-E), BERT (Bidirectional Encoder Representations from Transformers), and LaMDA (Language Model for Dialogue Applications).

[0094] In addition, the “chatbot” according to the present invention may be implemented to receive information that is difficult to obtain from multiple-choice questions in relation to the user’s indication. For example, the “chatbot” according to the present invention may include at least one of a rule-based chatbot, an artificial intelligence chatbot, a natural language processing chatbot, and a hybrid chatbot.

[0095] As illustrated in (b) of FIG. 3b, the control unit (130) can conduct an interactive survey (340) with the user's indications. More specifically, the control unit (130) can conduct an interactive survey (340) to collect information related to indications that are difficult to verify in detail through a selection survey.

[0096] For example, the control unit (130) may generate and provide interactive survey questions (341) related to the user's indications on a service page (1000) provided to the user terminal (10). In this case, the control unit (130) may provide the interactive survey (340) through at least one of a large language model (140) and a chatbot so that the user can freely respond to the interactive survey questions (341).

[0097] In the present invention, a process of receiving response data for a survey from a user terminal may be performed (S220, see FIG. 2).

[0098] The control unit (130) can receive selectable survey response data including a response matched to an item selected by user input among a plurality of selectable items from the user terminal (10). As illustrated in (a) of FIG. 3b, the control unit (130) can receive selectable response data including a response matched to an item selected by user input among a plurality of selectable items for a question related to the user's indication included in a specific selectable survey (331) (e.g., “Are you currently having difficulty with daily life or exercise due to pain?”).

[0099] As illustrated in (b) of FIG. 3b, the control unit (130) can receive conversational survey response data for conversational survey questions (340) from the user terminal (10). Specifically, the control unit (130) can provide at least one conversational survey question (341) to the user terminal (10) on the service page (1000) and receive the user's natural language input (342) for the conversational survey question (341) in a manner that converses with the user. Furthermore, the control unit (130) can process the natural language input (342) received from the user terminal (10) as input for at least one of a large language model (140) and a chatbot.

[0100] For example, the control unit (130) can process natural language input (342) entered in the conversational survey (340) as input to the large language model (140) and receive conversational survey response data from the large language model (140). As another example, the control unit (130) can process natural language input (342) entered in the conversational survey (340) as input to the chatbot and receive conversational survey response data from the chatbot.

[0101] Furthermore, the control unit (130) can provide a preset number of interactive survey questions to the user terminal (10) and receive the user's interactive survey response data for each of the interactive survey questions. At this time, the control unit (130) can provide a sample answer for the interactive survey question (341) along with the interactive survey question (341) so that important information regarding the interactive survey question (341) can be entered.

[0102] In the present invention, a process of evaluating a user's exercise performance ability from exercise motion data received from a user terminal using a previously trained motion evaluation model can be performed (S230, see FIG. 2).

[0103] The control unit (130) can collect exercise motion data of the user and evaluate the user's current exercise performance ability by using at least one of the camera and sensor unit provided in the user terminal (10). Here, “exercise motion data” may include at least one of an image and video of the user’s motion performed for evaluating exercise performance ability provided in the user terminal. For example, the control unit (130) can receive exercise motion data for a specific posture of the user captured by the camera provided in the user terminal (10).

[0104] Specifically, the control unit (130) may activate at least one sensor provided in the user terminal (10) to evaluate the user's exercise performance ability. For example, when an exercise performance ability evaluation is performed from the user terminal (10), the control unit (130) may activate a sensor unit including a microphone, a camera, and a plurality of different sensors provided in the user terminal (10).

[0105] Furthermore, the control unit (130) can receive sensing information based on the activation of at least one of the microphone, camera, and sensor unit provided in the user terminal (10). For example, the control unit (130) can receive exercise motion data including at least one of the user’s exercise motion image and exercise motion video from the camera based on the activation of at least one of the camera and microphone. At this time, the video received through the camera can be displayed in real time on the display unit of the user terminal (10).

[0106] Furthermore, the control unit (130) can analyze at least one of the user's static posture, joint range of motion, balance ability, and core strength using a previously learned motion evaluation model (132). Here, the motion evaluation model (132) is a posture estimation model learned using a learning data set containing position information for joint points, and may refer to an artificial intelligence model learned to estimate the user's (U) movement from movement data. Here, "joint point" may refer to a plurality of joints of the user (U) (or a part of the user's (U) body including joints). And, "key point" may refer to an area corresponding to each of the user's (U) plurality of joint points in at least one of an image and video containing the user's movement.

[0107] The control unit (130) can use a pre-learned motion evaluation model (132) to extract key points corresponding to the user's joint points from the user's exercise motion data, and evaluate the user's (U) exercise motion based on an analysis of the positional relationship between the extracted key points. In the present invention, a series of processes for evaluating the user's exercise performance ability from exercise motion data using key points extracted through the motion evaluation model (132) can be named the “exercise performance ability evaluation process.”

[0108] In the present invention, the physical space and subject where the exercise performance evaluation process takes place are not separately distinguished, and it can be described as taking place in the exercise program providing system (100). The exercise performance evaluation process can be performed using key points extracted from the motion evaluation model (132).

[0109] More specifically, the control unit (130) can analyze motion data received from the camera on a frame-by-frame basis to extract key points corresponding to joint points corresponding to the user's motion. For example, the motion evaluation model (132) may include various object detection algorithms. For example, the motion evaluation model (132) may include an algorithm that ensembles multiple bounding boxes (Weighted Box Fusion, WBF). However, it is obvious that the control unit (130) is not limited to the object detection algorithms described above and can use various object detection algorithms capable of detecting objects corresponding to the user (U) from images or videos.

[0110] In the present invention, the training data for which the operation evaluation model (132) performs training may be stored in a database (200), and a part of this database (200) may also be named a “training data DB.” Further details regarding the training data will be described later.

[0111] Furthermore, the control unit (130) can perform training for the motion evaluation model (132) based on a training data set existing in the training data DB. As previously described, the training data set may include location information of joint points matched with the user's joint points. The control unit (130) can train the motion evaluation model (132) to extract key points corresponding to the user's joint points from the exercise motion data using the training data set.

[0112] As illustrated in (a) of FIG. 4a, the control unit (130) can perform a motion analysis evaluation (410) for the user. Here, the “motion analysis evaluation (410)” may include at least one of a static posture evaluation (or body shape analysis) and a joint range of motion evaluation. Specifically, the control unit (130) can perform a motion analysis evaluation (410) for each of a plurality of different body parts to evaluate the user’s exercise performance ability at the user terminal (10).

[0113] For example, the control unit (130) can analyze the user's body shape from an image (or video) of the user's static posture captured from the user terminal (10) using a pre-learned motion evaluation model (132). Specifically, the control unit (130) can analyze the alignment status of the user's body parts (e.g., head, neck, shoulders, pelvis, knees, etc.) from the image (or video) of the static posture using the pre-learned motion evaluation model (132).

[0114] Furthermore, the control unit (130) may perform an evaluation of the joint range of motion for each of a plurality of different body parts in order to evaluate the user's exercise performance ability at the user terminal (10). Specifically, the control unit (130) may use a pre-learned motion evaluation model (132) to perform an analysis of at least one of the joint travel distance, joint movement speed (or acceleration), and the user's body balance, body equilibrium, and body alignment state (e.g., leg axis alignment state, spine alignment state, etc.) included in the exercise video to be analyzed.

[0115] For example, the control unit (130) can use a camera provided in the user terminal (10) to capture the user's movement based on a plurality of shooting topics for evaluating the range of motion of the joint. At this time, among the plurality of shooting topics (first shooting topic to fourth shooting topic), the first shooting topic (411) may be related to a “static posture,” the second shooting topic (412) may be related to a “range of motion of the joint (shoulder),” the third shooting topic (413) may be related to a “range of motion of the joint (elbow),” and the fourth shooting topic (414) may be related to a “range of motion of the joint (torso). Meanwhile, for each of the plurality of shooting items corresponding to the plurality of shooting topics (411 to 414), at least one of the shooting target scene (or scene), shooting method, and shooting order may be matched and exist.

[0116] The control unit (130) receives user exercise motion data according to the shooting method and shooting order matched to the shooting topic selected from the user terminal (10), and can perform user motion analysis evaluation (410) from the exercise motion data.

[0117] As illustrated in (b) and (c) of FIG. 4a, the control unit (130) can perform an evaluation of the user's balance ability and core strength (420, 430) using a pre-learned motion evaluation model (132). Specifically, the control unit (130) can evaluate at least one of the user's balance ability and core strength from the user's exercise motion data performing a pre-set exercise motion (e.g., one-leg standing motion, plank motion, squat motion, etc.). At this time, the control unit (130) can provide a sample video of the pre-set exercise motion to the user terminal (10) and induce the user to perform the pre-set exercise motion according to the provided sample video.

[0118] The control unit (130) measures the time a user maintains a preset exercise movement (e.g., one-leg standing posture, plank posture) in a proper posture and, based on the maintenance time, can perform at least one of an evaluation of the user's balance ability and an evaluation of core strength (420). For example, the control unit (130) can evaluate the measured maintenance time according to preset time intervals to perform an evaluation of the user's balance ability and core strength. Here, the “preset time intervals” may include multiple different time intervals. The control unit (130) can evaluate the user's balance ability and core strength as “insufficient (or low)” based on the first time interval (e.g., less than 1 minute) among the multiple different intervals, “average (or medium)” based on the second time interval (e.g., 1 minute to 3 minutes), and “good (or high)” based on the third time interval (e.g., more than 3 minutes). At this time, the pre-set time interval according to the present invention is not limited to the described examples and can be set in various ways by the exercise program providing system (100).

[0119] As another example, the control unit (130) measures the number of times a user performs a preset exercise movement (e.g., squat movement) in a proper posture for a preset time (e.g., 30 seconds), and based on the number of times, can perform at least one of the user's balance ability evaluation and core strength evaluation (420). For example, the control unit (130) can evaluate the measured number of times according to a preset number interval to perform the user's balance ability evaluation and core strength evaluation. Here, the “preset number interval” may include multiple different number intervals. The control unit (130) can evaluate the user’s balance ability and core strength as “insufficient (or low)” based on the first repetition interval (e.g., less than 10) among a plurality of different repetition intervals, “average (or medium)” based on the second repetition interval (e.g., 10 to 20), and “good (or high)” based on the third repetition interval (e.g., more than 20). At this time, the preset repetition interval according to the present invention is not limited to the described example and can be set in various ways by the exercise program providing system (100).

[0120] As illustrated in FIG. 4b, the control unit (130) can receive exercise motion data (440) by using at least one of the camera and sensor units provided in the user terminal (10). Furthermore, the control unit (130) can process the exercise motion data (440) as input to a pre-learned motion evaluation model (132) to generate an evaluation result (450) of the user's exercise performance ability.

[0121] Specifically, the exercise performance evaluation result (450) may include at least one of the user's motion analysis evaluation result (451), balance ability evaluation result (452), and core strength evaluation result (453). Furthermore, the control unit (130) may comprehensively evaluate the user's exercise performance based on the user's motion analysis evaluation result (451), balance ability evaluation result (452), and core strength evaluation result (453) based on pre-set evaluation criteria. Here, "pre-set evaluation criteria" may refer to criteria by which medical personnel evaluate the user's exercise performance based on the evaluation of the user's static posture, joint range of motion, balance ability, and core strength. The control unit (130) may generate an exercise performance evaluation result (450) that further includes a comprehensive evaluation result (454) for the comprehensive evaluation.

[0122] In the present invention, a process of generating a prompt to be input into a large language model can be carried out using at least one of the response data to the survey and the results of the exercise performance evaluation (S240, see FIG. 2).

[0123] As illustrated in FIG. 5, the control unit (130) can generate a prompt (520) to be input into a large language model (140) from survey response data (510), which includes multiple-choice survey response data (511) and interactive survey response data (512), and exercise performance evaluation results (450), using a prompt generation unit (131). Specifically, the prompt generation unit (131) can generate a prompt (520) that includes a response matched to an item selected by user input among the multiple selection items for a question related to the user's indication included in the multiple-choice survey. For example, the prompt generation unit (131) can generate a prompt (520) that includes information such as a specific pain area, degree of pain, and exercise equipment possessed in relation to the user's indication included in the multiple-choice survey response data (511).

[0124] Furthermore, the prompt generation unit (131) can generate a prompt (520) that includes the user's natural language input included in the interactive survey response data (512) and the answer of the large language model (140) to the natural language input.

[0125] Additionally, the prompt generation unit (131) can generate a prompt (520) that includes the user's motion analysis evaluation result (451) included in the user's exercise performance evaluation result (450). Specifically, the control unit (130) can identify a body part in which the user's joint range of motion satisfies a preset condition based on the motion analysis evaluation result (451). Here, the “preset condition” may refer to a condition satisfied based on the user’s joint range of motion not corresponding to a normal standard range or being limited to below a specific threshold. At this time, the “normal standard range” and the specific threshold may be set by at least one of the medical staff terminal (20) and the exercise program providing system (100).

[0126] Furthermore, the control unit (130) can generate a prompt to include exercise motion content related to a specific body part in the exercise program. As previously described, the “exercise motion content” according to the present invention may include various information such as the name of a specific exercise motion, the number of times the exercise is performed, the method of the exercise, and a video of the exercise motion. For example, the control unit (130) can generate a prompt to include exercise motion content related to the “waist (lumbar spine)” in the exercise program based on the fact that the joint range of motion of the “waist (lumbar spine)” in the motion analysis evaluation result (451) does not correspond to the normal standard range and satisfies a preset condition.

[0127] Additionally, the control unit (130) can generate a prompt (520) including a balance ability evaluation result (452) and a core muscle strength evaluation result (453). Furthermore, the control unit (130) can generate a prompt (520) including a comprehensive evaluation result (454) of the user's exercise performance ability.

[0128] In the present invention, a process of inputting a prompt into a large language model and generating an exercise program related to the user's indication through the large language model can be carried out (S250, see FIG. 2).

[0129] A control unit (130) according to the present invention can input a prompt (520) generated based on at least one of survey response data and exercise performance evaluation results into a large language model (140). The large language model (140) according to the present invention can generate an exercise program to be provided to a user based on the input prompt (520). Here, the “exercise program” is configured to include a plurality of exercise modules, each of which is matched to a different exercise type and may include at least one exercise motion content related to the exercise type matched to each exercise module. At this time, the different exercise types may mean at least one of a warm-up exercise type, a main exercise type, and a cool-down exercise type.

[0130] In the present invention, the fact that the large language model (140) generates an exercise program can also be understood as the control unit (130) generating an exercise program using the large language model (140). Below, the combination of the control unit (130) generating an exercise program using the large language model (140) and the large language model (140) generating an exercise program may be used interchangeably.

[0131] As illustrated in FIG. 6, the control unit (130) can extract multiple different exercise motion contents from a database (200) containing an exercise program DB (610) by using at least one of survey response data and exercise performance evaluation results based on prompt engineering. Here, prompt engineering may refer to a technique for designing and optimizing input text (prompt) to effectively utilize a natural language processing (NLP) model.

[0132] According to the present invention, the exercise program DB (610) may store and contain multiple different exercise motion contents (611 to 613) related to each of the various body parts (or pain parts) of a person. As previously described, the “exercise motion contents” may include various information such as a name for a specific exercise motion, exercise time information, exercise difficulty information, exercise type information, exercise body part information, exercise posture information, and exercise equipment information. At this time, each of the information (e.g., name for a specific exercise motion, exercise time information, exercise difficulty information, exercise type information, exercise body part information, exercise posture information, exercise equipment information) included in the multiple exercise motion contents (611 to 613) stored in the exercise program DB (610) may be labeled (or tagged) with a pre-set attribute value.

[0133] For example, based on exercise type information matched to a plurality of exercise motion contents (611 to 613) stored in the exercise program DB (610), a first type attribute value (e.g., “1”) may be labeled to the exercise motion content of the “warm-up exercise” type, and a second type attribute value (e.g., “2”) may be labeled to the exercise motion content of the “main exercise” type.

[0134] As another example, exercise difficulty information included in a plurality of exercise motion contents (611 to 613) stored in the exercise program DB (610) may be labeled with a preset attribute value based on the exercise type matched to the exercise motion content. Specifically, for exercise motion content matched to the “warm-up exercise” type, at least one of a first difficulty attribute value (e.g., “1”), a second difficulty attribute value (e.g., “2”), and a third difficulty attribute value may be labeled. On the other hand, for exercise motion content matched to the “main exercise” type, at least one of a first difficulty attribute value (e.g., “1”), a second difficulty attribute value (e.g., “2”), a third difficulty attribute value, a fourth difficulty attribute value (e.g., “4”), and a fifth difficulty attribute value (e.g., “5”) may be labeled. The description of the attribute values ​​described above is merely one example, and the attribute values ​​labeled in the information constituting the exercise motion content are not limited to the described examples and can be set in various ways.

[0135] Furthermore, the large language model (140) can extract attribute values ​​for each of the multiple pieces of information included in the exercise motion content in order to extract at least one exercise motion content to be included in the exercise program. The large language model (140) receives attribute values ​​corresponding to each of the extracted multiple pieces of information and can extract exercise motion content including information corresponding to the extracted attribute values ​​from the exercise program DB (610).

[0136] The control unit (130) can use a large language model (140) to extract multiple different exercise motion contents from the exercise program DB (610) and generate at least one exercise module based on an exercise type matched to each of the extracted multiple exercise motion contents. For example, the control unit (130) can identify exercise motion contents related to the user's pain area based on information regarding the user's pain area included in the prompt (520) among the multiple different exercise motion contents (611 to 613) from the exercise program DB (610). Additionally, the control unit (130) can identify exercise motion contents related to a body part where the user's joint range of motion satisfies a preset condition based on the user's motion analysis evaluation result included in the prompt (520) among the multiple different exercise motion contents (611 to 613) from the exercise program DB (610).

[0137] Furthermore, the control unit (130) may generate multiple different exercise modules based on the exercise type matched to each of the specified exercise motion contents. For example, the control unit (130) may generate exercise modules such that exercise motion contents of the “warm-up exercise” type among the specified exercise motion contents (e.g., “both hip flexion 1 (621)”, “right hip external rotation 1 (622)”) are included in the warm-up exercise module (620). As another example, the control unit (130) may generate exercise modules such that exercise motion contents of the “main exercise” type among the specified exercise motion contents (e.g., “leg straight raise 1 (631)”, “leg straight raise 2 (632)”) are included in the main exercise module (630). At this time, the control unit (130) can use a large language model (140) to generate an exercise program (600) composed of a warm-up exercise module (620), a main exercise module (630), and a cool-down exercise module (640) in that order.

[0138] Furthermore, the large language model (140) can generate an exercise program that allows a user to continuously perform the same exercise movements based on exercise posture information included in the exercise movement content, for the continuity of the movement posture in a specific exercise module. Specifically, the large language model (140) can extract multiple exercise movement contents containing the same exercise posture information and generate an exercise program in which the multiple exercise movement contents are arranged continuously for a predetermined number (e.g., 3).

[0139] Meanwhile, the control unit (130) can generate an exercise program including exercise motion content that can be performed using the exercise equipment (650) possessed by the user by using a large language model (140). Specifically, the large language model (140) can extract at least one exercise motion content including specific exercise equipment information by using the user's response data to a question related to exercise equipment from the survey response data included in the prompt (520). For example, the large language model (140) can extract at least one exercise motion content including exercise equipment information corresponding to the "foam roller" and "stretching band" based on the user's response in the survey response data that they have a "foam roller" and a "stretching band." Furthermore, the large language model (140) can generate an exercise program including the extracted at least one exercise motion content.

[0140] Furthermore, the control unit (130) may set the difficulty level (660) of the exercise program based on at least one of the survey response data and the exercise performance evaluation results. Here, the “difficulty level (660) of the exercise program” may refer to the intensity level of the exercise program provided in relation to the user’s indication. At this time, the difficulty level (660) of the exercise program according to the present invention may be set to any one of a plurality of pre-set levels (e.g., Grade A to Grade F). As an example, in the difficulty level (660) of the exercise program, “Grade A” may mean the lowest intensity and “Grade F” may mean the highest level.

[0141] More specifically, according to the present invention, for each of the plurality of grades, the difficulty level of the exercise movement content constituting each of the warm-up exercise module, the main exercise module, and the cool-down exercise module may be set and exist. For example, when the difficulty level (660) of the exercise program is set to “Grade C,” the warm-up exercise module and the cool-down exercise module may include exercise movement content in which the difficulty level of the exercise movement content includes at least one of “1,” “2,” and “3”. On the other hand, the main exercise module may include exercise movement content in which the difficulty level of the exercise movement content includes at least one of “2” and “3”.

[0142] In this way, according to the present invention, for each of the plurality of grades (e.g., Grade A to Grade F), the difficulty level of the exercise movement content constituting the warm-up exercise module, the main exercise module, and the cool-down exercise module may be set. In the present invention, the difficulty level of the exercise movement content included in different exercise modules according to the difficulty level set in the exercise program is not limited to the described examples and can be set in various ways.

[0143] Furthermore, the large language model (140) can generate an exercise program (600) based on the difficulty level (660) set in the exercise program, such that each different exercise module includes exercise movement content that satisfies a pre-set difficulty condition. Here, the “pre-set difficulty condition” may mean a condition in which the difficulty level of at least one exercise movement content included in a specific exercise module corresponds to the difficulty level of the exercise movement content according to the difficulty level grade set in the exercise program.

[0144] Meanwhile, the control unit (130) can use a large language model (140) to generate an exercise program corresponding to the total exercise time (670) that is pre-set in the medical staff terminal (20). More specifically, the large language model (140) can include multiple exercise motion contents in different exercise modules to correspond to the total exercise time that is pre-set, based on exercise time information included in the exercise motion contents. As an example, if the total exercise time that is pre-set in the medical staff terminal (20) is “20” minutes, the control unit (130) can set an exercise time of 5 minutes for the warm-up exercise module and the cool-down exercise module, respectively, and set an exercise time of 10 minutes for the main exercise module. Furthermore, the large language model (140) can combine multiple exercise motion contents to satisfy the exercise time set in each exercise module, thereby generating an exercise program that reflects the total exercise time (670) that is pre-set in the exercise program.

[0145] In the present invention, a process of transmitting a request for review of an exercise program generated through a large language model to a medical staff terminal may be carried out (S260, see FIG. 2).

[0146] As illustrated in FIG. 7, the control unit (130) may transmit a review request (710) for the exercise program (600) to the medical staff terminal (20) for the medical staff's approval of the generated exercise program (600). Here, the “review request” may mean a request to perform either approval or modification of the generated exercise program. The control unit (130) may provide a service page (1000) associated with the review request (710) to the medical staff terminal (20) to receive feedback from the medical staff regarding the review request (710).

[0147] Furthermore, the control unit (130) may display information (720) associated with the exercise program on the service page (1000). Here, the “information (720) associated with the exercise program” may include an exercise module list for the plurality of exercise modules constituting the exercise program. At this time, the exercise module list (not shown) may include at least one exercise motion content included in each of the plurality of exercise modules and at least one of a review icon indicating whether medical personnel have reviewed each of the plurality of exercise modules.

[0148] At this time, the medical staff can input medical staff feedback (730) through the service page (1000) provided on the medical staff terminal (20). Although the information (720) related to the exercise program and the medical staff feedback (730) shown in FIG. 7 are expressed as text, this should be understood as merely one embodiment. That is, it is not limited to the expression in FIG. 7, and the information (720) related to the exercise program and the medical staff feedback (730) may be provided in other formats (e.g., tables, images, lists, etc.). The information (720) related to the exercise program and the medical staff feedback (730) will be explained in more detail below with reference to the relevant drawings.

[0149] In the present invention, in response to a request for review, a process of receiving feedback on an exercise program from a medical staff terminal may be carried out (S270, see FIG. 2).

[0150] The control unit (130) can receive feedback (730) from medical staff regarding the exercise program through the service page (1000). Here, the feedback may be at least one of a first feedback approving the exercise program and a second feedback requesting modification of the exercise program. Specifically, the control unit (130) may approve the exercise program based on a review request provided to the medical staff terminal (20) upon the occurrence of an approval event for the exercise program.

[0151] Specifically, the “approval event” according to the present invention may be generated by the first feedback of the medical staff regarding the exercise program at the medical staff terminal (20) of the medical staff (D). Here, the “first feedback” may include at least one of a first approval input, which means user input for an approval icon provided on a service page associated with a review request provided on the medical staff terminal (20), and a second approval input based on natural language text input into the large language model (140). At this time, the “user input” may mean a selection input for a specific icon displayed on the service page. For example, the “user input” may be performed in at least one of a tap, double tap, long press, click, swipe, drag, pinch in, pinch out, and rotate.

[0152] The control unit (130) can specify a generated exercise program based on the occurrence of an approval event, and provide the specified exercise program as the final exercise program (740) to the user terminal (10). The specification and provision of an exercise program according to the present invention will be described in detail below with reference to the relevant drawings.

[0153] Meanwhile, the control unit (130) can modify the exercise program based on a review request provided to the medical staff terminal (20) based on the occurrence of a modification event for the exercise program. Specifically, the “modification event” according to the present invention may be generated by a second feedback from the medical staff regarding the exercise program at the medical staff terminal (20) of the medical staff (D). Here, the “second feedback” may include at least one of a first modification input, which refers to user input for a modification icon provided on a service page associated with a review request provided to the medical staff terminal (20), and a second modification input based on natural language text corresponding to a query from the medical staff input into the large language model (140). Furthermore, the control unit (130) can perform modification (or review) of the generated exercise program based on the large language model (140) based on the occurrence of the modification event.

[0154] As illustrated in FIG. 8a, the control unit (130) can perform modification (or review) of the exercise program through a service page associated with a request for review of the exercise program using a large language model (140). Specifically, the large language model (140) can perform question-and-answer with medical staff based on survey response data (510) containing at least one of patient history data (810), optional survey response data (511), and interactive survey response data (512) stored in the database (200), and exercise performance ability evaluation results (450). Here, the “patient history data (810)” may include various historical information related to the user’s indications, such as i) name, ii) date of birth, iii) type of indication of the user, iv) exercise history, v) past medical history, vi) prescription information, vii) history of response data to a survey, and viii) exercise prognosis information of multiple different users (or patients).

[0155] The control unit (130) can receive medical staff inquiries (821, 823) requesting modification (or review) of an exercise program on a service page provided to the medical staff terminal (20). Furthermore, the control unit (130) can generate a prompt corresponding to the medical staff inquiry (821) using a prompt generation unit (131), and process the generated prompt as input to a large language model (140) to generate an answer to the medical staff inquiry (821). The control unit (130) can perform question-and-answer with the medical staff by providing the generated answer to the medical staff inquiry (821) on a service page provided to the medical staff terminal (20).

[0156] In the present invention, a process of specifying an exercise program based on feedback can be carried out (S280, see FIG. 2).

[0157] In addition, the present invention may proceed with the process of providing a specific exercise program to the user terminal (S290, see FIG. 2).

[0158] As previously described, the control unit (130) may receive at least one of a first feedback approving the exercise program and a second feedback requesting modification of the exercise program from the medical staff terminal (20) in response to a request for review of the exercise program. At this time, the control unit (130) may perform different data processing processes for the exercise program according to the feedback received from the medical staff terminal (20). Specifically, the different data processing processes may include at least one of a first process of specifying the generated exercise program and providing the specific exercise program to the user terminal based on the first feedback received from the medical staff terminal, and a second process of modifying the exercise program based on the second feedback received from the medical staff terminal.

[0159] As previously explained, the control unit (130) can specify an exercise program as a final exercise program according to the first process based on an approval event generated from receiving first feedback from the medical staff terminal (20). Furthermore, the control unit (130) can provide the specified exercise program to the user terminal (10).

[0160] Meanwhile, the control unit (130) can modify the exercise program according to the second process based on a modification event that occurs from receiving second feedback from the medical staff terminal (20). For example, in the second process, a medical staff query related to at least one specific exercise module among a plurality of exercise modules constituting the exercise program can be received from the medical staff terminal (20) through a service page associated with a review request.

[0161] As illustrated in FIG. 8a, the control unit (130) may receive a medical staff inquiry (823) corresponding to a second feedback from the medical staff terminal (20) for modifying the exercise program. As an example, the control unit (130) may receive medical staff inquiries from the medical staff terminal (20) requesting modifications to the exercise program, such as a medical staff inquiry requesting to add specific exercise motion content to the warm-up exercise module and a medical staff inquiry requesting to modify the difficulty level of the main exercise module.

[0162] The control unit (130) can obtain information about a specific exercise module for which a medical staff query (823) has been received and process the medical staff query (823) as input to a large language model (140) to obtain an answer from the large language model (140) related to the specific exercise module. Furthermore, the control unit (130) can modify the specific exercise module using the answer from the large language model related to the specific exercise module.

[0163] As illustrated in (a) and (b) of FIG. 8b, the control unit (130) may display information associated with an exercise program on a service page associated with a review request. Here, the information associated with the exercise program may include an exercise module list (831) for the plurality of exercise modules constituting the exercise program. At this time, the exercise module list (831) may include at least one exercise motion content included in each of the plurality of exercise modules (620 to 640) and at least one of a review icon (832, 833) indicating whether a medical professional has reviewed each of the plurality of exercise modules.

[0164] Specifically, the control unit (130) may provide at least one candidate exercise motion content (841 to 843) associated with the specific exercise module (840) to be modified to the medical staff terminal (20) based on a medical staff query (834) related to the specific exercise module (840). Furthermore, the control unit (130) may identify, from the medical staff terminal (20), the specific exercise motion content included in the specific exercise module (840) and the specific candidate exercise motion content (842, 843) to be changed among the at least one candidate exercise motion content (841 to 843). For example, the control unit (130) may identify the specific candidate exercise motion content (842, 843) based on user input (850) regarding the motion change icon (844) displayed on the service page from the medical staff terminal (20).

[0165] The control unit (130) can receive a selection signal for the user input (850) and update a prompt to be input into the large language model (140). Specifically, the prompt generation unit (131) can generate an updated prompt in response to the selection signal so that specific candidate exercise action content (842, 843) is included in the specific exercise module (840). Furthermore, the control unit (130) can change the specific exercise action content included in the specific exercise module (840) to the specific candidate exercise action content (842, 843). Specifically, the control unit (130) can process the updated prompt as input to the large language model (140) to generate a modified exercise program through the large language model (140).

[0166] Meanwhile, the control unit (130) may suggest modifications to specific exercise motion content that utilizes exercise equipment not possessed by the user, based on exercise equipment information included in the survey response data from the medical staff terminal (20). As illustrated in FIG. 8c, the control unit (130) may use the survey response data to provide the medical staff terminal (20) with an exercise equipment list (650) among the information associated with the exercise program provided on the service page. At this time, the exercise equipment list (650) may include a plurality of exercise equipment (651 to 653) required for performing the exercise motion content included in the exercise program. Additionally, the exercise equipment list (650) may include possession icons (861, 862) indicating whether the user possesses each of the plurality of exercise equipment.

[0167] The control unit (130) may receive a medical staff inquiry (830) requesting a modification of exercise motion content that requires an exercise machine that the user does not possess among a plurality of exercise machines. As described above, the control unit (130) may use the medical staff inquiry received from the medical staff terminal to update a prompt and process the updated prompt as input to a large language model (140). Furthermore, the control unit (130) may use the large language model (140) to output a plurality of candidate exercise motion contents (871 to 873) corresponding to the answer (970) corresponding to the medical staff inquiry (830) to the medical staff terminal (20).

[0168] Furthermore, the control unit (130) can identify, from the medical staff terminal (20), a specific exercise motion content that requires an exercise device not possessed by the user among the plurality of candidate exercise motion contents (871 to 873) and a specific candidate exercise motion content (872, 873) to be changed. For example, the control unit (130) can identify specific candidate exercise motion contents (842, 843) based on user input (880) for an action change icon (874) displayed on a service page from the medical staff terminal (20).

[0169] The control unit (130) can receive a selection signal for the user input (880) and update a prompt to be input into the large language model (140). Specifically, the prompt generation unit (131) can generate an updated prompt in response to the selection signal, so that specific candidate exercise action content (872, 873) in the prompt is changed to specific exercise action content that requires exercise equipment not possessed by the user. Furthermore, the control unit (130) can change the specific exercise action content that requires exercise equipment not possessed by the user to specific candidate exercise action content (872, 873). Specifically, the control unit (130) can process the updated prompt as input to the large language model (140) to generate a modified exercise program through the large language model (140).

[0170] Furthermore, the control unit (130) can use a large language model to perform a modification request based on a medical staff query and provide the modified exercise program to the user terminal (10). Specifically, the control unit (130) can specify the modified exercise program and provide it to the medical staff terminal (20), and specify the modified exercise program as the final exercise program.

[0171] As illustrated in FIG. 9a, the control unit (130) can provide a response (911) to the medical staff terminal (20) regarding the result of the modification request. Furthermore, the control unit (130) can receive a medical staff query (912) requesting the creation of a modified exercise program. The control unit (130) can identify the modified exercise program as the final exercise program through a large language model (140).

[0172] As illustrated in FIG. 9b, the control unit (130) can provide information associated with the final exercise program (900) to the medical staff terminal (20). Specifically, the control unit (130) can provide a plurality of exercise modules (920, 930, 940, 950) constituting the final exercise program (900) to the medical staff terminal (920). At this time, each of the plurality of exercise modules (920, 930, 940, 950) may include modified exercise motion content (921, 922, 931, 932) based on a medical staff query. Furthermore, the control unit (130) can provide the final exercise program (900) including at least one of the difficulty level (960) and total exercise time information (970) of the final exercise program based on the modified exercise motion content (921, 922, 931, 932).

[0173] Furthermore, the control unit (130) may perform either a first process or a second process according to user input applied to the medical staff terminal (20) regarding at least one of the approval icon (981) and the modification icon (982) provided on the service page associated with the review request. For example, the control unit (130) may specify the modified exercise program as the final exercise program according to the first process in response to user input (990) regarding the approval icon (981) for the final exercise program (900) provided on the service page from the medical staff terminal (20). Furthermore, the control unit (130) may provide the specified exercise program to the user terminal (10).

[0174] The method and system for providing an exercise program using a large language model according to the present invention can generate a user-customized exercise program by utilizing response data to a survey provided to a user terminal and the results of an evaluation of the user's exercise performance ability. Through this, the present invention can generate an exercise program of a difficulty level that considers the user's current condition, thereby maximizing the rehabilitation effect for the user's indications.

[0175] Furthermore, the method and system for providing an exercise program using a large language model according to the present invention can provide an exercise program with guaranteed expertise and stability by utilizing a large language model and reflecting feedback from medical staff on the modules constituting the exercise program.

[0176] Furthermore, the method and system for providing an exercise program using a large language model according to the present invention can provide a generated exercise program to a user terminal based on feedback from medical staff. Through this, patients can receive rehabilitation treatment without having to visit a hospital located far away, thereby enabling easy access to rehabilitation treatment. Additionally, medical staff can conveniently monitor the patient's rehabilitation exercise through electronic devices and provide monitoring-based feedback to enhance the effectiveness of the patient's exercise therapy.

[0177] Meanwhile, computer-readable media include all types of recording devices in which data that can be read by a computer system is stored. Examples of computer-readable media include HDD (Hard Disk Drive), SSD (Solid State Disk), SSD (Silicon Disk Drive), ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc.

[0178] Furthermore, the computer-readable medium may be a server or cloud storage that includes a storage and is accessible to an electronic device via communication. In this case, the computer may download the program according to the present invention from the server or cloud storage via wired or wireless communication.

[0179] Furthermore, in the present invention, the computer described above is an electronic device equipped with a processor, namely a CPU (Central Processing Unit), and no special limitations are placed on its type.

[0180] Meanwhile, the above detailed description should not be interpreted restrictively in all respects but should be considered exemplary. The scope of the invention should be determined by a reasonable interpretation of the appended claims, and all modifications within the equivalent scope of the invention are included within the scope of the invention.

Claims

1. A step of providing at least one questionnaire related to the user's indication to the user terminal; A step of receiving response data for the survey from the user terminal; A step of evaluating the user's exercise performance ability from exercise motion data received from the user terminal using a previously trained motion evaluation model; A step of generating a prompt to be input into a large language model using at least one of the response data to the above survey and the results of the above exercise performance evaluation; A step of inputting the above prompt into the above large language model to generate an exercise program related to the user's indications through the above large language model; A step of transmitting a request for review of the exercise program generated through the above large language model to a medical staff terminal; In response to the above review request, a step of receiving feedback on the exercise program from the medical staff terminal; Based on the above feedback, a step of specifying the exercise program; and A method for providing an exercise program using a large language model, characterized by including the step of providing the aforementioned specified exercise program to the user terminal.

2. In Paragraph 1, The above exercise program is configured to include a plurality of exercise modules, and Each of the above plurality of motion modules is, A method for providing an exercise program using a large language model, characterized by being matched to different types of exercise and including at least one exercise motion content related to the type of exercise matched to each of the exercise modules.

3. In Paragraph 2, The step of receiving the above feedback is, A step of providing a service page linked to the review request to the medical staff terminal; A step of displaying information associated with the exercise program on the above service page; and A method for providing an exercise program using a large language model, characterized by including the step of receiving feedback from the medical staff regarding the exercise program through the service page.

4. In Paragraph 3, The information associated with the above exercise program is, It includes a list of exercise modules for the plurality of exercise modules constituting the above exercise program, and The above list of exercise modules is, A method for providing an exercise program using a large language model, characterized by including at least one exercise motion content included in each of the plurality of exercise modules and at least one review icon indicating whether the medical staff has reviewed each of the plurality of exercise modules.

5. In Paragraph 3, The above feedback is, A method for providing an exercise program using a large language model, characterized by having at least one of a first feedback approving the exercise program and a second feedback requesting modification of the exercise program.

6. In Paragraph 5, In the step of specifying the above exercise program, Performing different data processing processes for the exercise program according to the feedback received from the medical staff terminal, The above different data processing processes are, A first process for specifying the generated exercise program based on receiving the first feedback from the medical staff terminal and providing the specified exercise program to the user terminal, and A method for providing an exercise program using a large language model, characterized by including at least one of a second process for modifying the exercise program based on receiving the second feedback from the medical staff terminal.

7. In Paragraph 6, In the second process above, A step of receiving a medical staff query related to at least one specific exercise module among the plurality of exercise modules constituting the exercise program from the medical staff terminal through the above service page; A step of obtaining information regarding the specific exercise module in which the medical staff query is received, and processing the medical staff query as input to the large language model to obtain an answer from the large language model related to the specific exercise module; and A method for providing an exercise program using a large language model, characterized by including the step of modifying a specific exercise module using the answer of the large language model related to the specific exercise module.

8. In Paragraph 7, The step of modifying the specific motion module mentioned above is, A step of providing at least one candidate exercise motion content associated with the specific exercise module to be modified to the medical staff terminal based on the medical staff query related to the specific exercise module; A step of specifying, from the medical staff terminal, a specific exercise motion content included in the specific exercise module and a specific candidate exercise motion content to be changed among the at least one candidate exercise motion content; and A method for providing an exercise program using a large language model, characterized by including the step of changing specific exercise motion content included in the specific exercise module into specific candidate exercise motion content.

9. In Paragraph 2, The above different types of exercise are, A method for providing an exercise program using a large language model characterized by having at least one of a warm-up exercise type, a main exercise type, and a cool-down exercise type.

10. In Paragraph 1, The above at least one survey is, A multiple-choice questionnaire consisting of a question related to the indications of the user and multiple selection items corresponding to each of the multiple different responses to the question, and A method for providing an exercise program using a large language model, characterized by including at least one interactive questionnaire capable of receiving natural language input from the user terminal in relation to the indications of the user.

11. In Paragraph 10, The step of receiving response data for the above survey is, A step of receiving the selectable survey response data from the user terminal, the data including a response matched to an item selected by user input among the plurality of selectable items; and A method for providing an exercise program using a large language model, characterized by including the step of processing the natural language input entered in the above-mentioned interactive survey as input to the large language model and receiving the interactive survey response data from the large language model.

12. In Paragraph 1, The step of evaluating the exercise performance ability of the above user is, A step of receiving motion data for a specific posture of the user captured from a camera provided in the user terminal; and A method for providing an exercise program using a large language model, characterized by including the step of analyzing at least one of the user's static posture, joint range of motion, balance ability, and core strength using the previously learned motion evaluation model.

13. In Paragraph 12, The step of generating a prompt to be input into the above-mentioned large language model is, A step of specifying a body part in which the joint range of motion of the user satisfies a preset condition; and A method for providing an exercise program using a large language model, characterized by including the step of generating the prompt so that exercise motion content related to the specified body part is included in the exercise program.

14. A control unit that provides at least one survey related to the user's indication to a user terminal, and a communication unit that receives response data for the survey from the user terminal. The above control unit is, An exercise program providing system using a large language model, characterized by evaluating the user’s exercise performance ability from exercise motion data received from the user terminal using a previously trained motion evaluation model, generating a prompt to be input into a large language model using at least one of the response data to the survey and the exercise performance ability evaluation result, inputting the prompt into the large language model to generate an exercise program related to the user’s indication, transmitting a request for review of the exercise program generated through the large language model to a medical staff terminal, receiving feedback on the exercise program from the medical staff terminal in response to the review request, specifying the exercise program based on the feedback, and providing the specified exercise program to the user terminal.

15. A program that is executed by one or more processes in an electronic device and stored on a computer-readable recording medium, The above program is, A step of providing at least one questionnaire related to the user's indication to the user terminal; A step of receiving response data for the survey from the user terminal; A step of evaluating the user's exercise performance ability from exercise motion data received from the user terminal using a previously trained motion evaluation model; A step of generating a prompt to be input into a large language model using at least one of the response data to the above survey and the results of the above exercise performance evaluation; A step of inputting the above prompt into the above large language model to generate an exercise program related to the user's indications through the above large language model; A step of transmitting a request for review of the exercise program generated through the above large language model to a medical staff terminal; In response to the above review request, a step of receiving feedback on the exercise program from the medical staff terminal; Based on the above feedback, a step of specifying the exercise program; and A program stored on a computer-readable recording medium characterized by including instructions that perform the step of providing the above-mentioned specific exercise program to the user terminal.

Images