Method and apparatus for generating personalized avatar through multi-biosignal sensing

Abstract

The present invention relates to a method for generating a personalized avatar through multi-biosignal sensing, comprising the steps of: generating, through an initial data input of at least one from among body information, medical history, lifestyle habits and biosignals, an avatar that reflects the current appearance and health condition of a user; collecting, in real time, data and biosignals related to lifestyle patterns and the lifestyle habits and inputting the collected data into a trained artificial intelligence model so as to predict the real-time health condition of the user; and changing the appearance or behavior of the avatar in order to emphasize and express disease risk information according to the predicted real-time health condition of the user.

Description

Method and device for creating personalized avatars using multiple biosignal sensing

[0001] This invention was made under the support of the Ministry of Science and ICT, under the task identification number 2710037161 and task number 00357879. The research management specialized organization of the said task is the National IT Industry Promotion Agency, the research project name is "Digital Innovation Challenge Leading Technology Development", the research project title is "AI-based Bio-Data Fusion and Generation Technology for Intelligent Personalized Chronic Disease Management", the main organization is Hanyang University Industry-Academic Cooperation Foundation, and the research period is from April 1, 2024 to December 31, 2027.

[0002] The present invention relates to a method and device for generating a personalized avatar through multiple biosignal sensing, and more particularly, to a method and device for generating a personalized avatar through multiple biosignal sensing, which estimates a chronic disease indicator through multiple biosignal sensing and reflects the estimated chronic disease indicator in a personalized avatar.

[0003] A smart home is a technology that uses internet-connected home appliances or smart devices to control all systems within a home. With the emergence of smart homes and the advancement of related technologies, home appliances can now acquire various biometric signals and lifestyle patterns from smart devices in real time. This has led to a growing trend of attempts to utilize these data to analyze and predict users' health.

[0004] A personalized avatar is a digital character or graphic representation created to reflect a user's personal characteristics, preferences, or biometric information. Users can view their personalized avatar on the screen of a smartphone, tablet, desktop, or laptop, allowing them to intuitively identify their personal characteristics, preferences, or biometric information.

[0005] With the advent of smart homes and personalized avatars, there is a need for a method and device for creating personalized avatars through multi-biosignal sensing that can estimate chronic disease indicators through multi-biosignal sensing and reflect the estimated chronic disease indicators in a personalized avatar to intuitively check the user's health status and lifestyle patterns, and to predict and intuitively understand the user's future health status.

[0006] The present invention provides a method and device for creating a personalized avatar through multiple biosignal sensing, which estimates chronic disease indicators through multiple biosignal sensing and reflects the estimated chronic disease indicators in a personalized avatar, thereby enabling intuitive confirmation of a user's health status, lifestyle patterns, etc., and also predicting and intuitively understanding the user's future health status.

[0007] In order to achieve the above object, the present invention is characterized in that it comprises a method for generating a personalized avatar through multiple bio-signal sensing, the step of generating an avatar that reflects the current appearance and health status of a user through inputting at least one initial data of body information, medical history, lifestyle habits, and bio-signals; a step of collecting data and bio-signals related to lifestyle patterns and lifestyle habits in real time, and inputting the collected data into a learned artificial intelligence model to predict the real-time health status of the user; and a step of changing the appearance or behavior of the avatar to emphasize and express disease risk information according to the predicted real-time health status of the user.

[0008] Preferably, the step of predicting the real-time health status of the user can collect data and bio-signals related to lifestyle patterns and habits in real time using smart home devices and mobile devices.

[0009] Preferably, the smart home device may include at least one of a cube-shaped controller, a smart robot, a motion sensor, a smart bed, a smart refrigerator, a smart mirror, a food scanner, and a VR digital therapy device.

[0010] Preferably, the food scanner and smart refrigerator collect data related to meal patterns and eating habits, such as ingested nutrients and calories, the smart mirror collects data related to body shape changes, exercise patterns and exercise habits, the VR digital therapeutic device collects data related to mental illness, such as gaze information, content of interest, and object of interest, the cube-shaped controller collects frequency data related to specific diseases, such as the user's movement and hand tremors, the smart robot follows the user and takes pictures, thereby collecting the user's lifestyle patterns and habits, the motion sensor collects data related to posture, and the smart bed can collect data related to body shape, sleep patterns, sleep habits, and biosignals.

[0011] Preferably, the mobile device may include at least one of a smart ring, a smart watch, a smart phone, a patch-type continuous glucose monitoring device, and a portable bio-signal measuring device.

[0012] Preferably, the smart ring, the smart watch, the smart phone, and the portable bio-signal measuring device collect one or more of photoplethysmography (PPG), electrocardiogram (ECG), and heart rate variability (HRV), and the patch-type continuous glucose monitoring device can collect blood glucose changes in the body in real time.

[0013] Preferably, the step of predicting the real-time health status of the user may extract weights for at least one chronic disease among hypertension, diabetes, blood sugar disorder, hyperlipidemia, obesity, depression, and eating disorders.

[0014] Preferably, the step of changing the appearance or behavior of the avatar can change the appearance or behavior of the avatar by the weight for one or more extracted chronic diseases.

[0015] Preferably, the method may further include a step of training the artificial intelligence model by inputting data and bio-signals related to lifestyle patterns and lifestyle habits and outputting weights for at least one chronic disease among hypertension, diabetes, blood sugar disorder, hyperlipidemia, obesity, depression, and eating disorders.

[0016] Preferably, the artificial intelligence model may include a multimodal encoder that integrates and processes data and biosignals related to life patterns and lifestyle habits, a long-term memory bank that stores the output of the multimodal encoder as long-term memory information, a query transformer that queries and processes the long-term memory information stored in the long-term memory bank, and a large-scale language model (LLM) that processes the output of the query transformer.

[0017] In addition, the present invention is a device for generating a personalized avatar through multiple bio-signal sensing, comprising: an avatar generating unit for generating an avatar reflecting the current appearance and health status of a user through input of at least one initial data of body information, medical history, lifestyle habits, and bio-signals; a learning unit for training an artificial intelligence model by inputting data and bio-signals related to lifestyle patterns and lifestyle habits and outputting weights for at least one chronic disease of hypertension, diabetes, blood sugar disorder, hyperlipidemia, obesity, depression, and eating disorders; a prediction unit for collecting data and bio-signals related to lifestyle patterns and lifestyle habits in real time and inputting the collected data into the trained artificial intelligence model to predict the real-time health status of the user; and an avatar changing unit for changing the appearance or behavior of the avatar to emphasize and express disease risk information according to the predicted real-time health status of the user.

[0018] The present invention has the advantage of being able to provide an immediate, impactful notification and warning to an individual by intuitively and quantitatively expressing whether the individual's health condition is improving or worsening compared to an avatar of the individual's current appearance through real-time bio-signal monitoring and bio-signal prediction.

[0019] FIG. 1 illustrates a flowchart of a method for creating a personalized avatar through multiple biosignal sensing according to an embodiment of the present invention.

[0020] FIG. 2 illustrates a created or modified avatar according to a method for creating a personalized avatar through multiple biosignal sensing according to an embodiment of the present invention.

[0021] Figure 3 shows the application fields of a method for creating a personalized avatar through multiple biosignal sensing according to an embodiment of the present invention.

[0022] FIG. 4 is a schematic diagram of a method for creating a personalized avatar through multiple biosignal sensing according to an embodiment of the present invention.

[0023] FIG. 5 is a drawing for explaining a step of changing the appearance or behavior of an avatar according to an embodiment of the present invention.

[0024] Figure 6 shows the structure of an artificial intelligence model according to an embodiment of the present invention.

[0025] Fig. 7 shows a configuration diagram of a personalized avatar generation device through multiple biosignal sensing according to an embodiment of the present invention.

[0026] A method for generating a personalized avatar through multiple biosignal sensing,

[0027] A step of creating an avatar reflecting the user's current appearance and health status through inputting at least one of initial data from among body information, medical history, lifestyle habits, and biosignals;

[0028] A step of collecting data and bio-signals related to lifestyle patterns and habits in real time, and inputting the collected data into a learned artificial intelligence model to predict the user's real-time health status; and

[0029] A step of changing the appearance or behavior of the avatar to emphasize and express disease risk information according to the predicted real-time health status of the user;

[0030] A method comprising:

[0031] Hereinafter, the present invention will be described in detail with reference to the contents described in the attached drawings. However, the present invention is not limited or restricted by the exemplary embodiments. The same reference numerals in each drawing indicate components that perform substantially the same functions.

[0032] The purpose and effects of the present invention can be naturally understood or made clearer by the following description, and the purpose and effects of the present invention are not limited solely by the following description. Furthermore, in describing the present invention, if a detailed description of known technologies related to the present invention is deemed to unnecessarily obscure the gist of the present invention, such detailed description will be omitted.

[0033] The terminology used herein is merely used to describe specific embodiments and is not intended to limit the present invention. The singular expression includes the plural expression unless the context clearly indicates otherwise. In this application, it should be understood that the terms "comprise" or "have" indicate the presence of a feature, number, step, operation, component, part, or combination thereof described in the description of the invention, but do not preclude the possibility of the presence or addition of one or more other features, numbers, steps, operations, components, parts, or combinations thereof.

[0034] While terms like "first" and "second" may be used to describe various components, these components should not be limited by these terms. These terms are used solely to distinguish one component from another. For example, without departing from the scope of the present invention, a first component could be referred to as a "second component," and similarly, a second component could also be referred to as a "first component."

[0035] Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. Terms defined in commonly used dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and shall not be interpreted in an idealized or overly formal sense unless explicitly defined herein.

[0036] When interpreting components, even if there is no explicit description, it is interpreted as including the margin of error. When describing temporal relationships, for example, when temporal continuity is described with phrases such as "after," "following," "next to," or "before," this also includes cases where the relationship is not continuous, unless "immediately" or "directly" is used.

[0037] Hereinafter, the technical configuration of the present invention will be described in detail with reference to the attached drawings.

[0038] Fig. 1 illustrates a flowchart of a method for generating a personalized avatar through multiple biosignal sensing according to an embodiment of the present invention. Referring to Fig. 1, the method for generating a personalized avatar through multiple biosignal sensing may include a step of generating an avatar (S100), a step of predicting a user's real-time health status (S300), a step of changing the avatar's appearance or behavior (S500), and a step of training an artificial intelligence model (S700).

[0039] FIG. 2 illustrates a generated or modified avatar in a method for generating a personalized avatar through multiple bio-signal sensing according to an embodiment of the present invention. Referring to FIG. 2, the method for generating a personalized avatar through multiple bio-signal sensing can generate and output an avatar that reflects the user's current appearance and health status, and at the same time, can change the avatar by reflecting the user's future appearance and health status predicted based on the user's lifestyle patterns, lifestyle habits, and bio-signals collected and stored in real time, or can output the current avatar and the future avatar simultaneously. In addition, the method for generating a personalized avatar through multiple bio-signal sensing can output an avatar reflecting the user's current appearance and health status and an avatar reflecting the user's future appearance and health status, and can output the current lifestyle patterns, lifestyle habits, and bio-signals, and can output trends, statistics, and predicted values ​​of the lifestyle patterns, lifestyle habits, and bio-signals.

[0040] Figure 3 illustrates the application fields of a method for generating a personalized avatar through multi-biosignal sensing according to an embodiment of the present invention. Referring to Figure 3, the method for generating a personalized avatar through multi-biosignal sensing can acquire a user's lifestyle patterns, lifestyle habits, and biosignals in a smart home environment. The method for generating a personalized avatar through multi-biosignal sensing can acquire lifestyle patterns, lifestyle habits, and biosignals from smart home AI agents, smart home devices, and mobile devices.

[0041] Referring to Figure 3, the method for generating a personalized avatar through multi-biosignal sensing can serve as a data platform that integrates and manages real-time acquired healthcare data (EHR) and medical clinical data (EMR) collected from medical institutions. The method for generating a personalized avatar through multi-biosignal sensing can utilize the integrated and managed data for AI diagnosis, AI prescription, and AI patient analysis, thereby supporting personalized treatment rather than group-tailored treatment, and can also be utilized in clinical decision support systems (CDSS). Ultimately, the method for generating a personalized avatar through multi-biosignal sensing can provide an environment where users can prevent diseases or manage their health remotely in real time without having to visit a medical institution in person.

[0042] FIG. 4 schematically illustrates a method for generating a personalized avatar through multiple biosignal sensing according to an embodiment of the present invention. Referring to FIG. 4, the step (S100) of generating an avatar can generate an avatar that reflects the user's current appearance and health status through the input of at least one initial data item among body information, medical history, lifestyle habits, and biosignals.

[0043] The avatar creation step (S100) can create a personal avatar similar to the user's face and body condition by inputting initial data such as the user's current physical information (height, weight, chest / waist / hip circumference, etc.), medical history, lifestyle habits, and biometric information. The avatar creation step (S100) can create an avatar similar to the user's face by inputting a user's photo.

[0044] The step (S300) of predicting the user's real-time health status collects data and bio-signals related to lifestyle patterns and habits in real time, and inputs the collected data into a learned artificial intelligence model to predict the user's real-time health status. The step (S300) of predicting the user's real-time health status collects the user's lifestyle patterns and habits (food, exercise, sleep, posture, etc.) in real time through the user's smart home device or mobile device, and can also collect various bio-signals (electrocardiogram, photoplethysmography, movement, etc.). The step (S300) of predicting the user's real-time health status recognizes lifestyle patterns, lifestyle habits, and bio-signal patterns using the learned artificial intelligence model, and can predict the user's health status. At this time, depending on whether the predicted value of a specific lifestyle pattern, lifestyle habit, or bio-signal pattern is likely to fluctuate significantly, the expression method of the individual avatar may vary in the step (S500) of changing the avatar's appearance or behavior to emphasize and express the corresponding disease risk information.

[0045] The step (S300) of predicting the user's real-time health status can collect data and bio-signals related to lifestyle patterns and habits in real time using smart home devices and mobile devices.

[0046] The smart home device may include at least one of a cube-shaped controller, a smart robot, a motion sensor, a smart bed, a smart refrigerator, a smart mirror, a food scanner, and a VR digital therapy device.

[0047] Food scanners and smart refrigerators can collect data related to eating patterns and habits, such as consumed nutrients and calories. Food scanners can analyze the user's food intake by photographing the user's table or plate. Food scanners can analyze the type, amount, and calorie intake of food consumed by the user. Step S300, which predicts the user's real-time health status through the collected data related to eating patterns and habits, can predict the user's future health status. Smart refrigerators can collect information such as the user's primary food intake, food discarded, the frequency with which the refrigerator is opened, and eating habits.

[0048] Smart mirrors can collect data related to body shape changes, movement patterns, and exercise habits. They can analyze the user's facial expressions, complexion, and posture to gather data that can be used to analyze the user's psychological state and health.

[0049] VR digital therapy devices can collect data related to mental illness, such as gaze information, content of interest, and objects of interest. VR digital therapy devices can provide mental health and cognitive therapy content to patients with mental illnesses such as depression. By tracking a user's gaze, VR digital therapy devices can determine whether their mental state is anxious or stable. They can also analyze the user's mental state by analyzing the objects of interest within the content based on their gaze.

[0050] Cube-shaped controllers can collect frequency data related to a user's movements and specific medical conditions, such as hand tremors. Cube-shaped controllers are devices that control electronic devices by moving a cube. For example, if a cube-shaped controller functions as a TV remote control, moving the cube forward could change the channel up, and moving it backward could change the channel down. Moving the cube to the right could increase the volume, and moving it to the left could decrease it. These cube controllers could be particularly useful for people with limited fine motor skills, such as the elderly, dementia patients, and those with Parkinson's disease.

[0051] As symptoms worsen in patients with dementia, Parkinson's disease, and other conditions, slowed movements and tremors can worsen. The Cube Controller can detect these slowed movements and tremors through frequency changes. Since frequency changes vary depending on the specific disease, it can provide data that can predict the likelihood of developing a specific disease. Furthermore, the Cube Controller can predict the likelihood of a specific disease worsening or reversing.

[0052] Smart robots can follow users and film them, collecting information about their lifestyle patterns and habits. They can collect data that can be used to analyze users' facial expressions, posture, movement patterns, and habits. Because they continuously track users, smart robots can collect real-time data on their sitting, lying, and moving times.

[0053] Motion sensors can collect posture-related data. They can be installed in chairs, beds, and other devices. Motion sensors can collect data on the user's sitting and lying time. Motion sensors can analyze a user's posture, determining whether the load is properly distributed, whether the posture is slouched or upright. This data can be used to determine the risk of developing musculoskeletal disorders, or whether the condition is likely to worsen or improve.

[0054] Smart beds can collect data and biometric signals related to body type, sleep patterns, and sleep habits. Smart beds can collect data on the user's sleep cycle, patterns, smartphone usage, snoring, apnea, and sleep talking.

[0055] The mobile device may include at least one of a smart ring, a smart watch, a smart phone, a patch-type continuous glucose monitoring device, and a portable bio-signal meter.

[0056] Smart rings, smart watches, smart phones, and portable biometric monitors can collect one or more of photoplethysmography (PPG), electrocardiogram (ECG), and heart rate variability (HRV).

[0057] Photoplethysmography (PV) refers to changes in blood flow through the vestibule. PV provides information that can predict not only heart rate but also vascular conditions, such as blood density. Diabetes is a disease in which blood glucose levels rise abnormally because it cannot enter cells and be used as an energy source. High blood glucose levels lead to higher blood viscosity (density) compared to healthy individuals. Because morphological changes in PV occur depending on blood viscosity, it can be used to diagnose diabetes.

[0058] If the PPG value is predicted to increase, the avatar can be changed to a personalized form that can notify the user of the risk of increased blood sugar levels, along with the predicted blood sugar level in the future (e.g., an avatar that appears fatter in proportion to the current body information and PPG change). Furthermore, if the PPG value is predicted to decrease and there is concern about hypoglycemia, an avatar can be created that can notify the user of the risk of hypoglycemia (e.g., a dizzy appearance). Alternatively, an avatar can be created that guides the user on appropriate actions to take (e.g., blood sugar check guide, treatment guidance).

[0059] The electrocardiogram (ECG) is a signal that represents the electrical activity of the heart, and along with the photoplethysmography signal, it can also be used to predict the health of the heart and blood vessels. For example, the time difference between the R-peak signal of the ECG, which occurs during the systole of the heart, and the R-peak of the photoplethysmography signal measured at the peripheral blood vessels simultaneously (PTT, Pulse Transit Time) can be used to predict systolic blood pressure (SBP). In addition, arterial stiffness is used to assess the risk of cardiovascular disease or predict axial atherosclerotic cardiovascular disease. Pulse wave velocity (PWV) is predicted using PTT and blood flow path length, which can be used as an indicator of arterial stiffness.

[0060] Therefore, if blood pressure or arterial stiffness is predicted to increase, an avatar can be created that can inform of the risk along with the predicted blood pressure and arterial stiffness values ​​in the future (e.g., holding the chest area in proportion to the current body information and ECG, PPG changes, expressing the heart area in the chest in red to intuitively express the risk, or collapsing due to myocardial infarction). Alternatively, an avatar can be created that guides the appropriate actions that an individual should take (e.g., blood pressure check guide, etc.).

[0061] Smart rings, smartwatches, smartphones, and portable bio-signal measuring devices may further include inertial motion units (IMUs). Inertial sensors measure acceleration, direction, angular velocity, and gravity, and can extract movement patterns such as walking, running, sitting, and standing, and can detect abnormal movements such as falls. Therefore, if the variability of movement patterns is predicted to increase through inertial sensor signal data, an avatar can be created that can notify the risk of these changes along with the predicted variability in walking, running, sitting, and standing in the future (e.g., an image of a fall occurring in proportion to the current body information and the amount of change in the inertial sensor). Alternatively, an avatar can be created that guides appropriate actions that an individual should take (e.g., recommending wearing a smartwatch or ring that can effectively detect the risk of falling).

[0062] Heart Rate Variability (HRV), which is a signal that lists the heartbeat as a series of times, is a biosignal that can quantitatively analyze the autonomic nervous system. By continuously measuring the balance between the sympathetic and parasympathetic nerves, it can be used as an indicator to monitor the imbalance of the autonomic nervous system due to mental health. Therefore, if the imbalance of heart rate variability is predicted to increase, an avatar can be created that can notify the risk of this along with the future fluctuation value (or mental health indicator) (e.g., a lonely or depressed appearance proportional to the current physical information and the amount of change in heart rate variability). Or, an avatar can be created that guides an appropriate action that an individual should take (e.g., inducing counseling).

[0063] Patch-type continuous glucose monitoring devices can collect real-time data on blood sugar levels. Because they measure blood sugar levels in real time, they can measure changes in blood sugar levels before and after meals, and based on these trends, they can predict disease risk and health status.

[0064] The step (S300) of predicting the user's real-time health status can extract weights for at least one chronic disease among hypertension, diabetes, blood sugar disorder, hyperlipidemia, obesity, depression, and eating disorders.

[0065] Step (S300) of predicting a user's real-time health status does not predict health status based on a single input data set, but rather predicts health status by comprehensively analyzing multiple input data sets using an artificial intelligence model. Because the chronic diseases that can be identified differ depending on whether the input data is used singly or in combination, the present invention utilizes a large-language multimodal artificial intelligence model that comprehensively analyzes the input data to extract weights for chronic diseases.

[0066] The step of changing the appearance or behavior of the avatar (S500) can change the appearance or behavior of the avatar to emphasize and express disease risk information according to the predicted real-time health status of the user.

[0067] FIG. 5 is a diagram illustrating a step for changing the appearance or behavior of an avatar according to an embodiment of the present invention. Referring to FIG. 5, the step (S500) for changing the appearance or behavior of an avatar may change the appearance or behavior of the avatar by the weights of one or more extracted chronic diseases.

[0068] The step (S700) of training an artificial intelligence model may train the artificial intelligence model by inputting data and bio-signals related to lifestyle patterns and lifestyle habits, and outputting weights for at least one chronic disease among hypertension, diabetes, blood sugar disorder, hyperlipidemia, obesity, depression, and eating disorders.

[0069] Fig. 6 illustrates the structure of an artificial intelligence model according to an embodiment of the present invention. Referring to Fig. 6, the artificial intelligence model may include a multimodal encoder that integrates and processes data and biosignals related to life patterns and lifestyle habits, a long-term memory bank that stores the output of the multimodal encoder as long-term memory information, a query transformer that queries and processes the long-term memory information stored in the long-term memory bank, and a large-scale language model (LLM) that processes the output of the query transformer.

[0070] Fig. 7 illustrates a configuration diagram of a personalized avatar generation device (10) using multiple biosignal sensing according to an embodiment of the present invention. Referring to Fig. 7, the personalized avatar generation device (10) using multiple biosignal sensing may include an avatar generation unit (100), a prediction unit (300), an avatar change unit (500), and a learning unit (700).

[0071] The avatar generation unit (100) can generate an avatar reflecting the user's current appearance and health status by inputting at least one of the following initial data: physical information, medical history, lifestyle habits, and bio-signals. The avatar generation unit (100) can perform the step (S100) of generating the avatar described above.

[0072] The prediction unit (300) collects data and bio-signals related to lifestyle patterns and habits in real time, and inputs the collected data into a trained artificial intelligence model to predict the user's real-time health status. The prediction unit (300) can perform the step (S300) of predicting the user's real-time health status described above.

[0073] The avatar change unit (500) can change the appearance or behavior of the avatar to emphasize disease risk information based on the predicted real-time health status of the user. The avatar change unit (500) can perform the step (S500) of changing the appearance or behavior of the avatar described above.

[0074] The learning unit (700) can train an artificial intelligence model by inputting data and bio-signals related to lifestyle patterns and habits and outputting weights for at least one chronic disease among hypertension, diabetes, blood sugar disorder, hyperlipidemia, obesity, depression, and eating disorders. The learning unit (700) can perform the step (S700) of training the aforementioned artificial intelligence model.

[0075] While the present invention has been described in detail through representative examples above, those skilled in the art will understand that various modifications to the above-described embodiments are possible without departing from the scope of the present invention. Therefore, the scope of the present invention should not be limited to the described embodiments, but should be defined not only by the claims described below but also by all changes or modifications derived from the claims and equivalent concepts.

[0076] The present invention provides a method and device for creating a personalized avatar through multiple biosignal sensing, which estimates chronic disease indicators through multiple biosignal sensing and reflects the estimated chronic disease indicators in a personalized avatar, thereby enabling intuitive confirmation of a user's health status, lifestyle patterns, etc., and also predicting and intuitively understanding the user's future health status.

Claims

1. A method for creating a personalized avatar through multiple biosignal sensing, A step of creating an avatar reflecting the user's current appearance and health status through inputting at least one of initial data from among body information, medical history, lifestyle habits, and biosignals; A step of collecting data and bio-signals related to lifestyle patterns and habits in real time, and inputting the collected data into a learned artificial intelligence model to predict the user's real-time health status; and A step of changing the appearance or behavior of the avatar to emphasize and express disease risk information according to the predicted real-time health status of the user; A method comprising:

2. In paragraph 1, The step of predicting the real-time health status of the above user is: A method for collecting data and bio-signals related to lifestyle patterns and habits in real time using smart home devices and mobile devices.

3. In paragraph 2, The above smart home device, A method comprising at least one of a cube-shaped controller, a smart robot, a motion sensor, a smart bed, a smart refrigerator, a smart mirror, a food scanner, and a VR digital therapy device.

4. In paragraph 3, The food scanner and smart refrigerator collect data related to eating patterns and eating habits, such as ingested nutrients and calories, the smart mirror collects data related to body shape changes, exercise patterns and exercise habits, the VR digital therapeutic device collects data related to mental illness, such as gaze information, content of interest, and object of interest, the cube-shaped controller collects frequency data related to specific illnesses, such as the user's movements and hand tremors, the smart robot follows the user and takes pictures, thereby collecting the user's lifestyle patterns and habits, the motion sensor collects data related to posture, and the smart bed collects data related to body shape, sleep patterns, sleep habits, and biosignals.

5. In paragraph 2, The above mobile device, A method comprising at least one of a smart ring, a smart watch, a smart phone, a patch-type continuous glucose monitoring device, and a portable bio-signal measuring device.

6. In paragraph 5, A method wherein the smart ring, the smart watch, the smart phone, and the portable bio-signal measuring device collect one or more of photoplethysmography (PPG), electrocardiography (ECG), and heart rate variability (HRV), and the patch-type continuous glucose monitoring device collects blood glucose changes in the body in real time.

7. In paragraph 1, The step of predicting the real-time health status of the above user is: A method for extracting weights for at least one chronic disease among hypertension, diabetes, blood sugar disorder, hyperlipidemia, obesity, depression, and eating disorders.

8. In paragraph 7, The steps for changing the appearance or behavior of the above avatar are: A method for changing the appearance or behavior of the avatar by the weight of one or more extracted chronic diseases.

9. In paragraph 1, A method further comprising a step of training the artificial intelligence model by inputting data and bio-signals related to lifestyle patterns and lifestyle habits and outputting weights for at least one chronic disease among hypertension, diabetes, blood sugar disorder, hyperlipidemia, obesity, depression, and eating disorders.

10. In paragraph 9, The above artificial intelligence model is, A method comprising: a multimodal encoder for integrating and processing data and biosignals related to life patterns and lifestyle habits; a long-term memory bank for storing the output of the multimodal encoder as long-term memory information; a query transformer for querying and processing the long-term memory information stored in the long-term memory bank; and a large-scale language model (LLM) for processing the output of the query transformer.

11. As a device for creating a personalized avatar through multiple bio-signal sensing, An avatar generation unit that generates an avatar reflecting the user's current appearance and health status through input of at least one initial data among physical information, medical history, lifestyle habits, and bio-signals; A learning unit that trains the artificial intelligence model by inputting data and bio-signals related to lifestyle patterns and habits and outputting weights for at least one chronic disease among hypertension, diabetes, blood sugar disorder, hyperlipidemia, obesity, depression, and eating disorders; A prediction unit that collects data and bio-signals related to life patterns and habits in real time and inputs the collected data into a learned artificial intelligence model to predict the user's real-time health status; and An avatar changing unit that changes the appearance or behavior of the avatar to emphasize and express disease risk information according to the predicted real-time health status of the user; A device comprising:

Images