System

Abstract

A system is provided.SOLUTION: A system, comprising: means for generating a personalized care plan using a generated AI model; means for displaying the generated care plan; means for predicting future health needs using the generated AI model; means for notifying care staff of the predicted results and the care plan; and means for collecting feedback from the user and utilizing it to retrain the model.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The technology of the present disclosure relates to a system. [Background technology]

[0002] Patent document 1 discloses a persona chatbot control method performed by at least one processor, the method including the steps of receiving a user utterance, adding the user utterance to a prompt including an instruction sentence related to a description of the chatbot character, encoding the prompt, and inputting the encoded prompt into a language model to generate a chatbot utterance in response to the user utterance. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Publication No. 2022-180282 Summary of the Invention [Problem to be solved by the invention]

[0004] As Japan's aging society progresses, traditional elderly care management systems face challenges in providing appropriate care tailored to individual health conditions and lifestyles. They also face difficulties in predicting the future health needs of elderly people and taking early action. This can lead to a decline in the quality of care and a loss of quality of life for elderly people. [Means for solving the problem]

[0005] The present invention solves the above problem by providing a system that includes a means for collecting a user's health data, a means for preprocessing the collected health data, a means for generating an individualized care plan using a generative AI model based on the preprocessed data, a means for displaying the generated care plan, a means for predicting future health needs using the generative AI model, a means for notifying care staff of the prediction results and the care plan, and a means for collecting feedback from users and using the feedback to retrain the model.

[0006] By receiving user input and collecting data via a screen device, and using deep learning models, the generative AI model improves the accuracy and predictive power of personalized care plans and provides more targeted care, thereby improving the quality of life for seniors and improving the quality of care.

[0007] "Users" refer to the elderly people and their caregivers who use this system.

[0008] "Health data" refers to information such as a user's number of steps, heart rate, sleep data, medical records, dietary habits, exercise amount, and medication intake.

[0009] "Means of collection" refers to systems including smartwatches, fitness trackers, medical institution databases, and user input interfaces.

[0010] "Preprocessing means" refers to the process of removing noise from collected data and extracting necessary features.

[0011] "Generative AI model" refers to an artificial intelligence model that generates personalized care plans from a user's features and predicts future health needs.

[0012] "Individualized care plan" refers to customized care instructions and recommendations based on a user's health and lifestyle.

[0013] "Display means" refers to a screen device or display interface for showing the individualized care plan or prediction results to the user or care staff.

[0014] "Health needs prediction" refers to the process of predicting future health risks and care needs based on a user's current data.

[0015] "Means of notifying care staff" refers to a system that communicates care plans and prediction results to care staff as alerts or messages.

[0016] "Means for collecting feedback" refers to an interface that allows users to input their thoughts and suggestions for improvement and sends that information to a server.

[0017] "Model retraining" refers to the process of updating a generative AI model based on new data and feedback from users to improve its accuracy. [Brief explanation of the drawings]

[0018] [Figure 1] 1 is a conceptual diagram showing an example of the configuration of a data processing system according to a first embodiment. [Figure 2] 1 is a conceptual diagram showing an example of main functions of a data processing device and a smart device according to a first embodiment. [Figure 3] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a second embodiment. [Figure 4] FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and smart glasses according to a second embodiment. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a third embodiment. [Figure 6] FIG. 11 is a conceptual diagram showing an example of main functions of a data processing device and a headset-type terminal according to a third embodiment. [Figure 7] FIG. 10 is a conceptual diagram showing an example of the configuration of a data processing system according to a fourth embodiment. [Figure 8]FIG. 10 is a conceptual diagram showing an example of main functions of a data processing device and a robot according to a fourth embodiment. [Figure 9] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 10] 1 shows an emotion map onto which multiple emotions are mapped. [Figure 11] FIG. 3 is a sequence diagram showing a processing flow of the data processing system according to the first embodiment. [Figure 12] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 1. [Figure 13] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system according to the second embodiment when an emotion engine is combined. [Figure 14] FIG. 10 is a sequence diagram showing the flow of processing in the data processing system in Application Example 2 when an emotion engine is combined. DETAILED DESCRIPTION OF THE INVENTION

[0019] An example of an embodiment of a system according to the technology of the present disclosure will be described below with reference to the accompanying drawings.

[0020] First, the terms used in the following description will be explained.

[0021] In the following embodiments, a coded processor (hereinafter simply referred to as a "processor") may be a single arithmetic device or a combination of multiple arithmetic devices. Furthermore, a processor may be a single type of arithmetic device or a combination of multiple types of arithmetic devices. Examples of arithmetic devices include a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), a GPGPU (General-Purpose computing on Graphics Processing Units), and an APU (Accelerated Processing Unit).

[0022] In the following embodiments, a coded RAM (Random Access Memory) is a memory in which information is temporarily stored and is used as a working memory by a processor.

[0023] In the following embodiments, the coded storage is one or more non-volatile storage devices that store various programs, various parameters, etc. Examples of non-volatile storage devices include flash memory (SSD (Solid State Drive)), magnetic disks (e.g., hard disks), and magnetic tapes.

[0024] In the following embodiments, a communication I / F (Interface) with a symbol is an interface including a communication processor, an antenna, etc. The communication I / F controls communication between multiple computers. Examples of communication standards applied to the communication I / F include wireless communication standards including 5G (5th Generation Mobile Communication System), Wi-Fi (registered trademark), Bluetooth (registered trademark), etc.

[0025] In the following embodiments, "A and / or B" is synonymous with "at least one of A and B." In other words, "A and / or B" means that it may be only A, only B, or a combination of A and B. Furthermore, in this specification, the same concept as "A and / or B" is also applied when three or more things are expressed connected by "and / or."

[0026] [First embodiment]

[0027] FIG. 1 shows an example of the configuration of a data processing system 10 according to the first embodiment.

[0028] 1, a data processing system 10 includes a data processing device 12 and a smart device 14. An example of the data processing device 12 is a server.

[0029] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0030] The smart device 14 includes a computer 36, a reception device 38, an output device 40, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The reception device 38, the output device 40, and the camera 42 are also connected to the bus 52.

[0031] The reception device 38 includes a touch panel 38A, a microphone 38B, and the like, and receives user input. The touch panel 38A detects contact with an indicator (for example, a pen or a finger) to receive user input by the touch of the indicator. The microphone 38B detects the user's voice to receive user input by voice. The control unit 46A transmits data indicating the user input received by the touch panel 38A and the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the data indicating the user input.

[0032] The output device 40 includes a display 40A and a speaker 40B, and presents data to the user 20 by outputting the data in a form of expression that the user 20 can perceive (for example, audio and / or text). The display 40A displays visible information such as text and images in accordance with instructions from the processor 46. The speaker 40B outputs audio in accordance with instructions from the processor 46. The camera 42 is a compact digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor.

[0033] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54.

[0034] FIG. 2 shows an example of the main functions of the data processing device 12 and the smart device 14.

[0035] 2, in the data processing device 12, a specific process is performed by the processor 28. A specific processing program 56 is stored in the storage 32. The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific process is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0036] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0037] In the smart device 14, the processor 46 performs the reception output process. The storage 50 stores a reception output program 60. The reception output program 60 is used in conjunction with the specific processing program 56 by the data processing system 10. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0038] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0039] MODE FOR CARRYING OUT THE INVENTION

[0040] The present invention is a system for creating personalized care plans and predicting health needs using generative AI in Japan's elderly care service management system. Specific embodiments of the system are described below.

[0041] System Configuration

[0042] This system consists of users (elderly people or their caregivers), a server, and terminals (smartphones, tablets, PCs, etc.).

[0043] Program processing flow

[0044] 1. Data Collection

[0045] server

[0046] The server periodically collects step counts, heart rate, and sleep data from smartwatches and fitness trackers worn by users.

[0047] It connects with medical institution databases to obtain users' medical records and health checkup results.

[0048] It also collects daily activity logs, such as dietary habits, exercise, and medication intake, which are entered by the user via the device.

[0049] 2. Data Preprocessing

[0050] server

[0051] Noise is removed from the collected data and missing values ​​are corrected.

[0052] The total number of steps taken each day is calculated based on the step count data from the smartwatch, and the resting heart rate is calculated based on the heart rate data.

[0053] Text data (e.g., meal details) is converted into structured data using natural language processing technology.

[0054] 3. Model Building

[0055] server

[0056] A generative AI model is trained using large amounts of pre-collected elderly care and medical data, and the model uses deep learning techniques.

[0057] The model generates personalized care plans and predicts future health needs based on the user's features.

[0058] 4. Generate personalized care plans

[0059] User

[0060] Users enter basic information (age, gender, medical history) and daily activity data through the device.

[0061] The server uses this information to generate an individualized care plan using a generative AI model.

[0062] The generated care plan is displayed on the device, allowing the user to check daily activity plans and recommended health behaviors.

[0063] 5. Anticipating health needs

[0064] server

[0065] The server uses a generative AI model to predict future health risks (e.g., risk of developing nutritional deficiencies or high blood pressure) based on the user's current health and lifestyle data.

[0066] The prediction results are communicated to the user and care staff in real time.

[0067] 6. Collaboration with care staff

[0068] Terminal

[0069] The individualized care plan and health needs prediction results received from the server are sent to the care staff's terminal.

[0070] The care staff adjusts specific care responses for the user based on the information received.

[0071] 7. Feedback Loops

[0072] User

[0073] The user provides feedback to the server via their device about the results of the care plan, their impressions, and changes in their physical condition.

[0074] The server analyzes the collected feedback and uses it to retrain the generative AI model.

[0075] Specific examples

[0076] Example 1: Data collection and preprocessing

[0077] The server automatically collects nighttime sleep data and daytime step count data from the smartwatch worn by the user every day, and cleanses this data to extract useful features.

[0078] Example 2: Generating personalized care plans

[0079] When a user enters their regular health check data into the terminal, the server uses that data to generate an individualized care plan, such as "walking 8,000 steps per day" or "light aerobic exercise twice a week," and displays it to the user.

[0080] Example 3: Predicting health needs

[0081] The server analyzes data entered by the user, such as a recent loss of appetite, and evaluates the estimated risk of nutritional deficiency. The generative AI model then suggests taking a multivitamin supplement and sends an alert to care staff.

[0082] The system allows users to implement personalized care plans to maintain and improve their health, and it also allows care staff to respond more accurately, improving the quality of elderly care.

[0083] The processing flow will be explained below.

[0084] Step 1: Data collection

[0085] server

[0086] The server works in conjunction with smartwatches and fitness trackers to periodically collect the user's steps, heart rate, and sleep data.

[0087] It connects to medical institution databases via API to obtain users' past medical records and regular health checkup results.

[0088] Users input their daily activities and dietary information into the device, and this data is also sent to the server.

[0089] Step 2: Data Preprocessing

[0090] server

[0091] The server cleanses the collected data, corrects missing and outlier values, and removes inappropriate data.

[0092] Standardize the data and extract features (e.g., average daily steps, resting heart rate, total calorie intake).

[0093] Text data (meal details, user feedback) is converted into structured data using natural language processing technology.

[0094] Step 3: Model Building

[0095] server

[0096] The server trains generative AI models using large datasets for elderly care.

[0097] Deep learning algorithms are used for training to build models that generate personalized care plans from user features and predict health needs.

[0098] Evaluate the accuracy of the model and tune the parameters as needed.

[0099] Step 4: Generate an individualized care plan

[0100] User

[0101] Users enter their basic information (age, gender, medical history) and daily activity data through the device.

[0102] server

[0103] The server collects the information submitted by the user and generates a personalized care plan using a generative AI model.

[0104] The generated care plan includes daily activity instructions and recommended health behaviors (e.g., aiming for 5,000 steps per day, eating a balanced diet).

[0105] Step 5: View Care Plan

[0106] Terminal

[0107] The designated care plan is displayed on the device screen, allowing the user to check the day's activities and health management instructions.

[0108] Step 6: Anticipate health needs

[0109] server

[0110] By analyzing a user's current health and lifestyle data, a generative AI model predicts future health risks (e.g., nutritional deficiencies, risk of chronic diseases).

[0111] Based on the prediction results, the system suggests necessary health management actions to the user (e.g., taking specific supplements, undergoing regular medical checkups).

[0112] Step 7: Working with notifications

[0113] Terminal

[0114] The prediction results and individualized care plan received from the server are sent to the care staff's terminal.

[0115] Care staff

[0116] Based on the information received, the care staff plans and implements specific care responses for the user.

[0117] Step 8: Gather feedback

[0118] User

[0119] Users input the results of the care plan, their impressions, and changes in their physical condition through their terminals, and provide feedback to the server.

[0120] server

[0121] The collected feedback is analyzed and used to improve and retrain the generative AI model.

[0122] Through feedback loops, the accuracy and effectiveness of the model and the system as a whole are continually improved.

[0123] Example 1

[0124] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0125] To improve the quality of elderly care, it is important to accurately generate individualized care plans and predict future health risks. However, conventional systems have difficulty centrally managing and effectively analyzing a user's diverse health data. Furthermore, they lack sufficient collaboration with care staff, making it difficult to provide appropriate care. Furthermore, they lack a mechanism for appropriately incorporating user feedback and updating the model, making it difficult to respond to ever-changing health conditions.

[0126] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0127] In this invention, the server includes: means for collecting health data from a user's activity tracking device; means for acquiring medical data in cooperation with a medical institution's database; means for collecting daily activity data entered by the user; means for performing noise reduction and missing value correction on the collected health data; means for calculating the total number of steps per day based on step count data and the resting heart rate based on heart rate data; means for converting text data into structured data using natural language processing technology; means for training a generative AI model using elderly care data and medical data; means for generating an individualized care plan using the generative AI model; means for displaying the generated care plan; means for predicting future health needs using the generative AI model; means for notifying care staff of the prediction results and the care plan; means for collecting user feedback and using it to retrain the model; means for care staff to adjust care responses to the user based on the care plan and prediction results received; and means for the user to provide feedback. This enables the generation of individualized care plans and prediction of health risks, improving the quality of elderly care. Furthermore, by appropriately incorporating user feedback and updating the model, care can be provided that is always tailored to the latest health status.

[0128] "User" refers to the elderly person or their caregiver who uses the system.

[0129] An "activity measuring device" is a device for collecting a user's physical activity data, and includes smart watches, fitness trackers, etc.

[0130] "Health data" refers to information such as a user's number of steps, heart rate, sleep data, medical records, health checkup results, dietary habits, exercise amount, and medication intake.

[0131] "Server" refers to the computer system that collects, preprocesses, analyzes, and stores user data, and trains and runs generative AI models.

[0132] "Noise reduction" refers to the process of removing unnecessary or erroneous information from collected data.

[0133] "Missing value correction" refers to the process of making predictions or inferences to fill in missing values ​​in data.

[0134] "Natural language processing technology" refers to technology for converting text data into a format that is easy for machines to understand.

[0135] "Generative AI model" refers to an AI (artificial intelligence) model that uses large amounts of data to generate personalized care plans and predict future health needs.

[0136] "Individualized care plan" refers to a care plan created based on a user's specific health condition and lifestyle habits.

[0137] "Health needs prediction" refers to the process of predicting future health risks and care needs based on a user's current data.

[0138] "Care staff" refers to professional people who provide care for older people.

[0139] "Feedback" refers to information provided by the user regarding the results of implementing the care plan, their impressions, and changes in their physical condition.

[0140] "Model retraining" refers to the process of updating a generative AI model with newly collected data and feedback to improve its accuracy.

[0141] This invention is a system for creating personalized care plans and predicting health needs using a generative AI model in an elderly care service management system in Japan. The system consists of users (elderly people or their caregivers), a server, and devices (smartphones, tablets, personal computers, etc.).

[0142] System Configuration

[0143] The system is configured as follows:

[0144] server

[0145] The server is responsible for:

[0146] 1. Periodically collect step count, heart rate, and sleep data from the user's activity tracking device (smartwatch or fitness tracker).

[0147] 2. Link with medical institution databases to obtain users' medical records and health checkup results.

[0148] 3. It also collects daily activity logs, such as dietary details, exercise volume, and medication intake, entered by the user through the device.

[0149] 4. Remove noise from the collected data and correct missing values.

[0150] 5. Calculate the total number of steps taken each day based on the step count data from the smartwatch, and calculate the resting heart rate based on the heart rate data.

[0151] 6. Convert text data (e.g., meal details) into structured data using natural language processing technology.

[0152] 7. Train generative AI models using elderly care and medical data, which use deep learning techniques.

[0153] 8. The model generates a personalized care plan and predicts future health needs based on the user's features.

[0154] Terminal

[0155] The device is used by users and care staff and has the following functions:

[0156] 1. The user enters basic information (age, gender, medical history) and daily activity data through the device.

[0157] 2. The device displays the personalized care plan generated using the generative AI model.

[0158] 3. Notify users and care staff in real time of health risk prediction results and the implementation results of care plans.

[0159] 4. Care staff will coordinate specific care responses for the user based on the information received.

[0160] User

[0161] Users provide daily activity data and feedback to the system:

[0162] 1. Users send health data to a server via their smartwatch or fitness tracker.

[0163] 2. The user enters information such as dietary habits and medication intake into the terminal application.

[0164] 3. Provide feedback on the results of the care plan, impressions, and changes in physical condition via the device application.

[0165] Specific examples

[0166] Example 1: Data collection and preprocessing

[0167] The server automatically collects nighttime sleep data and daytime step count data from the smartwatch worn by the user every day, and cleanses this data to extract useful features.

[0168] Example 2: Generating personalized care plans

[0169] When a user enters their regular health check data into the terminal, the server uses that data to generate an individualized care plan, such as "walking 8,000 steps per day" or "light aerobic exercise twice a week," and displays it to the user.

[0170] Example 3: Predicting health needs

[0171] The server analyzes data entered by the user, such as a recent loss of appetite, and assesses the estimated risk of nutritional deficiency. The generative AI model then suggests taking a multivitamin supplement and sends an alert to care staff. This system allows users to implement personalized care plans to maintain and improve their health. It also allows care staff to respond more accurately, improving the quality of elderly care.

[0172] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0173] Step 1: Data collection

[0174] Server collects smartwatch data

[0175] Input: Step count, heart rate, and sleep data from the user's smartwatch

[0176] Data processing: Collect data from the smartwatch periodically and store it in a database

[0177] Output: Raw health data

[0178] Specific operation: When the user syncs the smartwatch every morning, the server receives the night's sleep data and the previous day's step count data and stores them in a database.

[0179] Acquisition of medical data by the server

[0180] Input: Medical records and health check results from medical institution databases

[0181] Data processing: Data is acquired through the medical institution's API and saved as user health information.

[0182] Output: Retrieved medical data

[0183] Specific operation: Outside of the medical institution's scheduled maintenance period, the server sends an API request to obtain the user's latest blood test results and diagnostic data.

[0184] Manual data collection by the server

[0185] Input: Dietary information, exercise amount, and medication intake status entered by the user via the device

[0186] Data processing: Collect input data in real time and save it in a database

[0187] Output: Daily activity data

[0188] Specific operation: After dinner, the user uses the terminal application to input the details of their meal and medication intake, and the data is sent to the server in real time.

[0189] Step 2: Data Preprocessing

[0190] Server-based noise removal and missing value correction

[0191] Input: Collected health data

[0192] Data processing: data denoising and missing value imputation

[0193] Output: Cleaned health data

[0194] Specific operation: The server cleanses the data obtained from the smartwatch and fills in missing daily step count data with data from the previous day.

[0195] Server recalculation and structuring of data

[0196] Input: Data after noise removal and correction

[0197] Data processing: Counting step count data, calculating resting heart rate, structuring text data

[0198] Output: Recalculated and structured data

[0199] Specific operation: The server aggregates the step count data and calculates the total number of steps taken for the day. It also analyzes the text of meal descriptions entered by the user, such as "salad and grilled chicken," and extracts information about calories and nutrients as structured data.

[0200] Step 3: Model Building

[0201] Server-based training of generative AI models

[0202] Input: Elderly care data, medical data

[0203] Data processing: Using deep learning techniques to train AI models

[0204] Output: A trained generative AI model

[0205] How it works: Generative AI models are trained using historical medical datasets and iteratively trained until error is minimized.

[0206] Step 4: Generate an individualized care plan

[0207] User enters basic information

[0208] Input: User's basic information (age, gender, medical history), daily activity data

[0209] Data processing: Enter basic information into the system and send it to the server

[0210] Output: User basic information data

[0211] Specific operation: A user logs in to the application and enters their age, gender, and medical history on the profile screen.

[0212] Server-based care plan generation

[0213] Input: User's basic information and daily activity data

[0214] Data processing: Generative AI models are used to generate personalized care plans

[0215] Output: Individualized Care Plan

[0216] Specific operation: The server calls up an AI model based on the basic information entered and generates a daily activity schedule, recommended exercises, and a meal plan.

[0217] Displaying care plans on devices

[0218] Input: Generated personalized care plan

[0219] Data processing: Push notification of care plan to device

[0220] Output: The care plan displayed to the user

[0221] What it does: The care plan is pushed to the device, and when the user opens the app, a detailed activity plan is displayed on the dashboard.

[0222] Step 5: Anticipate health needs

[0223] Server-based health risk prediction

[0224] Input: User's current health and lifestyle data

[0225] Data processing: Predicting future health risks using generative AI models

[0226] Output: Health risk prediction results

[0227] How it works: Analyzes recent dietary data to assess the user's risk of vitamin deficiency. If the deficiency persists, the AI ​​model predicts the risk and generates an alert.

[0228] Real-time notifications

[0229] Input: Health risk prediction results

[0230] Data processing: Real-time notification of prediction results

[0231] Output: Notification to user and care staff

[0232] Specific operation: A push notification is sent to the care staff's device, and an alert is displayed stating, "User A is at risk of high blood pressure."

[0233] Step 6: Collaborate with care staff

[0234] Sending care information by device

[0235] Input: Personalized care plan and health risk prediction results received from the server

[0236] Data processing: Sending information to care staff's terminal

[0237] Output: Care plan and health risk prediction information

[0238] Specific operation: Information from the server is sent to an app dedicated to care staff, who can check it in real time.

[0239] Care staff coordination

[0240] Input: Care plan and health risk prediction information

[0241] Data processing: Coordination of care responses

[0242] Output: Adjusted care plan

[0243] Specific operation: During regular visits, care staff use information obtained from the server to fine-tune the user's care plan.

[0244] Step 7: Feedback Loop

[0245] User feedback

[0246] Input: Care plan implementation results, impressions, changes in physical condition

[0247] Data processing: Input the feedback into the system and send it to the server

[0248] Output: User feedback information

[0249] Specific operation: When the user enters feedback into the app, such as "I've been walking for a while, and as a result, I feel better," that information is sent to the server.

[0250] Server data reuse

[0251] Input: Collected feedback data

[0252] Data processing: Retraining generative AI models

[0253] Output: Improved generative AI model

[0254] Specific operation: The server analyzes the collected feedback data and performs re-training to improve the accuracy of the AI ​​model.

[0255] (Application example 1)

[0256] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0257] In conventional elderly care services, collecting health data and creating care plans is often done manually, making it difficult to provide personalized care. Furthermore, there is a lack of means to monitor health conditions in real time and quickly respond as needed. This poses a particular challenge when an elderly person becomes ill while traveling or when emergency care is required.

[0258] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0259] In this invention, the server includes means for collecting user health data, means for preprocessing the collected health data, means for generating an individualized care plan based on the preprocessed data using a generative AI model, means for displaying the generated care plan in the vehicle, means for predicting future health needs using the generative AI model, means for notifying a caregiver of the prediction results and the care plan, means for collecting feedback from the user and using it to retrain the model, means for optimizing a driving route according to the user's health condition, and means for detecting and notifying abnormal health data in real time, thereby enabling the provision of an individualized care plan, real-time health condition monitoring, and appropriate responses even while traveling.

[0260] "User" refers to an individual about whom health data is collected, particularly the elderly and those in need of health management.

[0261] "Health data" refers to information that indicates an individual's health status, such as heart rate, number of steps, sleep data, dietary content, amount of exercise, and medication intake.

[0262] "Preprocessing" refers to the processing of collected health data, such as noise removal, missing value correction, and data structuring.

[0263] A "generative AI model" is an artificial intelligence model trained using deep learning techniques to generate personalized care plans based on collected health data and predict future health needs.

[0264] A "personalized care plan" is a specific action plan created based on the user's individual health data and lifestyle, with the aim of maintaining or improving the health of a specific user.

[0265] An "autonomous vehicle" is a vehicle that is capable of driving independently and is equipped with the functionality to collect and process health data.

[0266] "Display means" refers to a device or function for visually showing the generated care plan to the user or caregiver (e.g., an in-car display or a smartphone).

[0267] "Health needs" are the activities and care requirements necessary to maintain or improve the user's future health status.

[0268] "Care aides" are professionals and caregivers who provide care to older people and people with health care needs.

[0269] "Notification means" refers to a device or system that notifies caregivers of the generated care plan and the predicted health needs (e.g., a smartphone app or an in-car system).

[0270] "Feedback" is information that the user reports, such as the results of implementing the care plan, their impressions, and changes in their physical condition.

[0271] "Model retraining" is a machine learning process that uses collected feedback data to further improve a generative AI model.

[0272] "Route optimization" is the process of selecting the optimal driving route and rest points based on the user's health condition.

[0273] "Real-time detection" is a function that can immediately recognize and respond to abnormalities in health data when they occur.

[0274] The present invention relates to a system for providing personalized health care in autonomous vehicles, particularly designed for the elderly and people with health care needs. Specific embodiments of the system are described below.

[0275] System Configuration

[0276] This system consists of a user (such as an elderly person), a server (the autonomous vehicle's infotainment system), a device (such as a smartwatch or fitness tracker) for collecting health data, and a caregiver (such as a caregiver or medical staff).

[0277] Program processing flow

[0278] Data collection

[0279] The server collects real-time health data such as heart rate, step count, and sleep data from the smartwatch or fitness tracker worn by the user. It also connects to medical institution databases to obtain the user's medical records and health checkup results. It also collects dietary information and daily activity logs entered by the user through a terminal inside the autonomous vehicle.

[0280] Data Preprocessing

[0281] The server removes noise from the collected health data and corrects missing values. It calculates the total number of steps taken each day based on the step count data from the smartwatch and calculates the resting heart rate based on the heart rate data. Text data such as meal details is converted into structured data using natural language processing technology.

[0282] Model Building

[0283] The server uses large amounts of pre-collected elderly care and medical data to train a generative AI model, which uses deep learning techniques to generate personalized care plans and predict future health needs based on the user's features.

[0284] Personalized care plan generation

[0285] Users enter basic information (age, gender, medical history) and daily activity data through the infotainment system. The server uses this information to generate a personalized care plan using a generative AI model. The generated care plan is displayed on a display inside the autonomous vehicle, allowing users to check their daily activity plan and recommended health behaviors.

[0286] Anticipating health needs

[0287] The server uses a generative AI model to predict future health risks (e.g., risk of developing nutritional deficiencies or high blood pressure) based on the user's current health and lifestyle data. The prediction results are notified to the user and caregivers in real time.

[0288] Cooperation with care staff

[0289] The personalized care plan and health needs prediction results received from the server are sent to the caregiver's device, who then adjusts specific care responses for the user based on the received information.

[0290] Feedback Loop

[0291] Users provide feedback to the server via their devices about the results of implementing the care plan, their impressions, and changes in their physical condition. The server analyzes the collected feedback and uses it to retrain the generative AI model.

[0292] Hardware and software used

[0293] Smartwatch: Collects heart rate, steps, and sleep data.

[0294] Autonomous vehicles: Infotainment systems for collecting and processing health data.

[0295] Server: Preprocessing data, training and running generative AI models.

[0296] Deep learning framework (TensorFlow or PyTorch): Implementing and training generative AI models.

[0297] Natural language processing technologies (SpaCy, NLTK, etc.): Structuring text data.

[0298] Specific examples

[0299] For example, if a user's heart rate increases while riding in an autonomous vehicle, the autonomous vehicle's system will collect heart rate data in real time and provide a care plan to "take a deep breath and relax" and a notification recommending a break at the next service area.

[0300] Prompt Sentence Examples

[0301] "The system monitors passengers' health status in real time based on heart rate and step count data during the ride, and immediately proposes an appropriate care plan if an abnormality is detected. It also suggests optimal routes and rest points depending on the passenger's condition."

[0302] The system enables personalized care plans, real-time health monitoring and appropriate responses, even while on the move.

[0303] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0304] Step 1:

[0305] A user wears a smartwatch or fitness tracker and gets into an autonomous vehicle. The device collects real-time heart rate, step count, and sleep data from the smartwatch and sends them to a server. The input of this step is the health data from the smartwatch, and the output is the raw data sent to the server.

[0306] Step 2:

[0307] The server preprocesses the received health data. Specifically, it removes noise and corrects missing values. For example, it complements missing data based on past data and removes obviously erroneous data. The input is the raw data collected in step 1, and the output is the cleansed health data.

[0308] Step 3:

[0309] The server uses a generative AI model to generate a personalized care plan based on the preprocessed data. For example, it may recommend deep breathing if the heart rate is high or light exercise if the step count is low. The input for this step is the preprocessed health data, and the output is a personalized care plan.

[0310] Step 4:

[0311] The generated care plan is presented to the user through the autonomous vehicle's infotainment system, where the user can visually review it and follow the instructions. The input is the generated personalized care plan, and the output is the care plan displayed to the user.

[0312] Step 5:

[0313] The server uses the generative AI model to predict the user's future health needs. For example, it evaluates the risk of rising blood pressure several months from now based on current data. The input of this step is the preprocessed health data, and the output is the predicted health risks and future health needs.

[0314] Step 6:

[0315] The server notifies the caregiver of the prediction results and the care plan. The caregiver can then adjust specific care responses for the user based on this information. The input is the prediction results and the care plan, and the output is a notification sent to the caregiver.

[0316] Step 7:

[0317] The server provides an interface for collecting feedback from users. Users input the results of implementing the care plan, their impressions, and changes in their physical condition. The input is the user's feedback, and the output is the feedback data stored on the server.

[0318] Step 8:

[0319] The server analyzes the collected feedback and uses it to retrain the generative AI model, which enables the model to provide more accurate care plans. The input is the feedback data, and the output is the retrained generative AI model.

[0320] Step 9:

[0321] The server optimizes the driving route according to the user's health condition. For example, it suggests appropriate rest areas according to the user's fatigue level. The input is health data detected in real time, and the output is the optimized driving route.

[0322] Step 10:

[0323] The server detects abnormal health data in real time and immediately notifies the user and caregivers, for example, sending an alert if the heart rate is abnormally high. The input is the health data collected in real time, and the output is the alert notification sent.

[0324] Furthermore, an emotion engine that estimates the user's emotion may be combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59 and perform identification processing using the user's emotion.

[0325] MODE FOR CARRYING OUT THE INVENTION

[0326] The present invention is a system for creating personalized care plans and predicting health needs using generative AI and an emotion engine in an elderly care service management system in Japan. Specific embodiments of the system are described below.

[0327] System Configuration

[0328] This system consists of users (elderly people or their caregivers), a server, terminals (smartphones, tablets, PCs, etc.), and an emotion engine.

[0329] Program processing flow

[0330] 1. Data Collection

[0331] server

[0332] The server periodically collects step counts, heart rate, and sleep data from smartwatches and fitness trackers worn by users.

[0333] It connects with medical institution databases to obtain users' medical records and health checkup results.

[0334] It also collects daily activity logs, such as dietary habits, exercise, and medication intake, which are entered by the user via the device.

[0335] 2. Collecting Emotional Data

[0336] Emotion Engine

[0337] The emotion engine uses facial recognition technology to analyze the user's facial expressions and assess their emotional state (e.g., happy, sad, or angry).

[0338] Analyze the tone of the user's voice to recognize their emotional state.

[0339] 3. Data Preprocessing

[0340] server

[0341] Noise is removed from the collected health and emotion data, missing and outliers are corrected, and inappropriate data is removed.

[0342] The data is standardized and features (e.g., average daily steps, resting heart rate, total calorie intake, and emotional data) are extracted.

[0343] Text data (meal details, user feedback) is converted into structured data using natural language processing technology.

[0344] 4. Model Building

[0345] server

[0346] A generative AI model is trained using a deep learning algorithm on a large dataset for elderly care.

[0347] The model builds a model that generates an individualized care plan based on the user's features and a model that predicts health needs.

[0348] It also takes into account sentiment data, evaluates the accuracy of the model, and tunes parameters as needed.

[0349] 5. Generate personalized care plans

[0350] User

[0351] Users enter basic information (age, gender, medical history) and daily activity data through the device.

[0352] server

[0353] The server collects the emotion data from the emotion engine and the information submitted by the user, and generates a personalized care plan using a generative AI model.

[0354] The generated care plan includes daily activity instructions and recommended health behaviors (e.g., aiming for 5,000 steps per day, eating a balanced diet).

[0355] 6. Care plan display

[0356] Terminal

[0357] The care plan is displayed on the device screen, allowing users to check their activities and health management instructions for the day.

[0358] 7. Anticipating health needs

[0359] server

[0360] By analyzing a user's current health, lifestyle, and emotional data, a generative AI model predicts future health risks (e.g., risk of nutritional deficiencies and high blood pressure).

[0361] Based on the prediction results, the system suggests necessary health management actions to the user (e.g., taking specific supplements, undergoing regular medical checkups).

[0362] 8. Notifications and Collaboration

[0363] Terminal

[0364] The prediction results and individualized care plan received from the server are sent to the care staff's terminal.

[0365] Care staff

[0366] Based on the information received, the care staff plans and implements specific care responses for the user.

[0367] 9. Feedback Collection

[0368] User

[0369] Users input the results of the care plan, their impressions, and changes in their physical condition through their terminals, and provide feedback to the server.

[0370] server

[0371] The collected feedback is analyzed and used to improve and retrain the generative AI model.

[0372] Through feedback loops, the accuracy and effectiveness of the model and the system as a whole are continually improved.

[0373] Specific examples

[0374] Example 1: Data collection and preprocessing

[0375] The user collects daily step counts and heart rate data from the smartwatch, and the server cleans the data to extract useful features. At the same time, the emotion engine collects the user's emotional data using the device's camera and microphone.

[0376] Example 2: Generating personalized care plans

[0377] Users who tend to be isolated input data from regular health checkups and daily life into the device, and the server generates an individualized care plan based on that data and emotional data obtained from the emotion engine, including recommendations such as "walking 7,000 steps a day" and "communicating with friends three times a week," and displays it to the user.

[0378] Example 3: Predicting health needs

[0379] The server analyzes data showing that the user enjoys walking but has recently been feeling depressed, and the generative AI model predicts that the user has a tendency toward seasonal depression. Reflecting the data from the emotion engine, the server recommends the user engage in daily light exercise and sun exposure, and notifies care staff to contact the user regularly.

[0380] The system allows users to receive scientifically-backed, personalized care plans and responses tailored to their emotional state, helping them maintain and improve their physical and mental health, while providing caregivers with more accurate information to provide specific care.

[0381] The processing flow will be explained below.

[0382] Step 1: Data collection

[0383] server

[0384] The server works in conjunction with smartwatches and fitness trackers to periodically collect the user's steps, heart rate, and sleep data.

[0385] It connects to medical institution databases via API to obtain users' past medical records and regular health checkup results.

[0386] Daily activity logs, such as what the user eats, how much exercise they do, and what medications they take, are also sent to the server.

[0387] Step 2: Collecting emotion data

[0388] Emotion Engine

[0389] The emotion engine uses the device's camera to capture the user's facial expressions and analyzes their emotional state (happiness, sadness, anger, etc.) using facial recognition technology.

[0390] When a user speaks using the device, the tone of their voice is analyzed through a microphone to recognize their emotional state.

[0391] Step 3: Data Preprocessing

[0392] server

[0393] The server cleanses the collected health and emotion data and corrects missing or outlier values.

[0394] The data is standardized and features (e.g., average daily steps, resting heart rate, total calorie intake, and emotional data) are extracted.

[0395] Text data (meal details, user feedback) is converted into structured data using natural language processing technology.

[0396] Step 4: Model Building

[0397] server

[0398] A generative AI model is trained using a large dataset for elderly care, using deep learning algorithms.

[0399] The model will be constructed to generate an individualized care plan based on the user's features, and to predict health needs.

[0400] The accuracy of the model is evaluated taking into account sentiment data and parameters are adjusted as needed.

[0401] Step 5: Generate an individualized care plan

[0402] User

[0403] Users enter their basic information (age, gender, medical history) and daily activity data into the device.

[0404] server

[0405] The server collects the emotion data from the emotion engine and the information submitted by the user, and generates a personalized care plan using a generative AI model.

[0406] The user's plan includes daily activity instructions and recommended health behaviors (e.g., aiming for 5,000 steps per day, adjusting diet, and taking mental health measures).

[0407] Step 6: View Care Plan

[0408] Terminal

[0409] The generated care plan is displayed on the terminal screen, and the user can check the day's activities and health management instructions.

[0410] Step 7: Anticipate health needs

[0411] server

[0412] The server analyzes the user's current health, lifestyle, and emotional data, and uses a generative AI model to predict future health risks (e.g., risk of nutritional deficiencies and high blood pressure).

[0413] Based on the prediction results, the system suggests necessary health management actions to the user (e.g., taking specific supplements, undergoing regular medical checkups).

[0414] Step 8: Working with notifications

[0415] Terminal

[0416] The prediction results and individualized care plan received from the server are sent to the care staff's terminal.

[0417] Care staff

[0418] Based on the information received, the care staff plans and implements specific care responses for the user.

[0419] Step 9: Gather feedback

[0420] User

[0421] The user inputs the results of the care plan, their impressions, and any changes in their physical condition into the terminal and provides feedback to the server.

[0422] server

[0423] The collected feedback is analyzed and used to improve and retrain the generative AI model.

[0424] Through feedback loops, the accuracy and effectiveness of the model and the system as a whole are continually improved.

[0425] Example 2

[0426] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0427] In elderly care services, it is challenging to provide care plans that take into account the user's individual health and emotional state. While conventional systems focus on collecting and analyzing health data, incorporating the user's emotional data is needed to provide more accurate personalized care plans. However, such systems face challenges due to the complexity of data collection and the need to assess the user's emotional state in real time. Furthermore, a means of clearly communicating the generated care plan and health risk predictions to the user is also required, and a mechanism for seamlessly aggregating subsequent feedback and retraining the model is lacking.

[0428] The identification processing by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting user health data, means for preprocessing the collected health data, means for generating an individualized care plan using a generative AI model based on the preprocessed data, means for collecting user emotion data using facial recognition technology and voice analysis technology, means for preprocessing the emotion data, means for generating an individualized care plan using the preprocessed emotion data and health data, means for displaying the generated care plan, means for predicting future health needs using the generative AI model, means for notifying care staff of the prediction results and the care plan, and means for collecting feedback from the user and using it to retrain the model. This makes it possible to provide an accurate individualized care plan based on the user's health condition and emotional state, and is expected to improve the accuracy of health risk predictions and the quality of care.

[0429] A "user" is an entity that utilizes the system to provide health and emotional data and receive a personalized care plan.

[0430] "Health data" refers to data collected by users in their daily lives, such as the number of steps taken, heart rate, sleep data, dietary content, amount of exercise, and medication intake.

[0431] "Emotional data" refers to data relating to the emotional state of a user obtained from facial expressions and tone of voice collected using face recognition technology and voice analysis technology.

[0432] "Facial recognition technology" is a technology that uses a camera to analyze a user's facial expressions and assess their emotional state.

[0433] "Voice analysis technology" is a technology that uses a microphone to analyze the tone and intonation of a user's voice to assess their emotional state.

[0434] A "generative AI model" is an algorithmic model that uses collected health and emotional data to generate personalized care plans and predict future health needs.

[0435] A "personalized care plan" is a plan that includes health care instructions and activity recommendations that are optimal for a user, generated based on the user's health and emotional data.

[0436] "Preprocessing" refers to processing of collected data, such as noise removal, missing value correction, outlier removal, data standardization, and feature extraction.

[0437] "Feedback" is information provided by the user regarding the results of implementing the care plan, their impressions, and changes in their physical condition, and is used to retrain the model.

[0438] "Care staff" are people who plan and implement specific care responses for users based on the individualized care plans and health needs prediction results generated using the system.

[0439] "Notification" refers to the act of notifying the user or care staff of the generated care plan or prediction results via a terminal.

[0440] MODE FOR CARRYING OUT THE INVENTION

[0441] This invention is a system for creating personalized care plans and predicting health needs in elderly care services using generative AI and an emotion engine. This system consists of a user, a server, a terminal, and an emotion engine.

[0442] 1. Users

[0443] Users wear a smartwatch or fitness tracker to record their daily activity data. They also input their daily activity log, including their diet, exercise, and medication intake, via a device (smartphone, tablet, PC, etc.). Furthermore, the device's camera and microphone are used to record data on their emotional state.

[0444] 2. Server

[0445] The server does the following:

[0446] Data collection: The server periodically collects data from the user's smartwatch or fitness tracker and connects with medical institution databases to obtain medical records and health checkup results.

[0447] Emotional data collection: We collect user emotional data using facial recognition and voice analysis technologies.

[0448] Data preprocessing: Noise removal, missing value correction, outlier removal, standardization, feature extraction, and text data structuring are performed on the collected health and emotion data.

[0449] Model Building: Train generative AI models using large-scale elderly care datasets to build and evaluate models that generate personalized care plans and predict health needs.

[0450] Prediction and generation: Using a generative AI model, future health risks are predicted based on the user's features and emotional data, and an individualized care plan is generated.

[0451] Notification and collaboration: Prediction results and care plans are notified to care staff devices.

[0452] 3. Terminal

[0453] The terminal has the following features:

[0454] Feedback collection: Users input the results of the care plan, their impressions, and changes in their physical condition via their terminal and provide feedback to the server.

[0455] Care plan display: The generated care plan is displayed on the terminal screen, allowing the user to check the health care instructions for the day.

[0456] Notification: Display the prediction results and care plan sent from the server and notify the care staff.

[0457] Specific examples

[0458] Example 1: Data collection and preprocessing

[0459] The user records their daily steps and heart rate on their smartwatch, and the server cleans and extracts features from the data. The emotion engine collects the user's emotion data using the device's camera and microphone.

[0460] Example 2: Generating personalized care plans

[0461] Users who tend to be isolated input data from regular health checkups and daily life into the device, and the server generates an individualized care plan based on that data and emotional data obtained from the emotion engine, including recommendations such as "walking 7,000 steps a day" and "communicating with friends three times a week," and displays it to the user.

[0462] Example 3: Predicting health needs

[0463] The server analyzes data showing that the user enjoys walking but has recently been feeling depressed, and the generative AI model predicts that the user has a tendency toward seasonal depression. Reflecting the data from the emotion engine, the server recommends the user engage in daily light exercise and sun exposure, and notifies care staff to contact the user regularly.

[0464] Prompt Sentence Examples

[0465] "The user is 70 years old, their primary daily activity is walking, and the emotion engine has detected their recent emotional state as 'sad'. Please generate a personalized care plan for this user."

[0466] The flow of the identification process in the second embodiment will be described with reference to FIG.

[0467] System program processing flow

[0468] Step 1: Data collection

[0469] The server periodically collects data such as step counts, heart rate, and sleep data from the smartwatch or fitness tracker worn by the user. As input, it receives biometric data from the smartwatch or fitness tracker. Specifically, it makes API calls from the smart device and stores the data in the server's database.

[0470] The server connects to the medical institution's database to retrieve the user's medical records and health checkup results. It receives data obtained from the medical institution's API as input. Specifically, it calls the API on a regular schedule, retrieves the necessary data, and stores it in the server's database.

[0471] Users input their daily activity logs, including their diet, exercise, and medication intake, via a device. The system receives inputs from users' manual entries and log data written on a daily basis. Specifically, data is entered through a dedicated application on the device and sent to a server. The output is the collected health data and daily activity log.

[0472] Step 2: Collecting emotion data

[0473] The emotion engine uses facial recognition technology to analyze video data acquired from the device's camera and analyze the user's facial expressions. It receives the video data acquired from the camera as input. Specifically, it uses a facial recognition algorithm to analyze the facial expressions and evaluate the user's emotional state, such as "happiness" or "sadness." The output is data indicating the emotional state.

[0474] The emotion engine analyzes audio data recorded using a microphone and recognizes emotions based on the user's tone of voice. It receives audio data from the microphone as input, analyzes the audio data using a voice analysis algorithm, and evaluates the user's emotional state. The output is audio data that indicates the user's emotional state.

[0475] Step 3: Data Preprocessing

[0476] The server performs noise removal on the collected health and emotion data. As input, it receives the collected raw data. Specifically, it detects outliers and missing values ​​and removes or corrects them. The output is the cleansed data.

[0477] The server normalizes the data and extracts features such as average daily steps, resting heart rate, total calorie intake, and emotional data. It receives the cleansed data as input and applies data normalization and feature extraction algorithms to convert it into analyzable data. The output is feature data for analysis.

[0478] The server converts text data (meal details, user feedback) into structured data using natural language processing technology. It receives text data entered by the user as input. Specifically, it applies a text analysis algorithm to convert the data into structured data, such as meal items. The output is structured text data.

[0479] Step 4: Model Building

[0480] The server trains a generative AI model using a large-scale dataset for elderly care. It receives a large amount of training data as input. Specifically, it uses a deep learning algorithm (e.g., TensorFlow or PyTorch) to train the model. The output is a trained generative AI model.

[0481] The server builds a model that generates an individualized care plan and a model that predicts health needs based on the user's features. It receives feature data as input. Specifically, it uses the generative AI model to create an optimal care plan and predict health risks for each user. The output is an individualized care plan generation model and a health needs prediction model.

[0482] Step 5: Generate an individualized care plan

[0483] The user inputs basic information (age, gender, medical history) and daily activity data through the terminal. The system receives basic information and daily activity data from the user as input. Specifically, the data is entered through an input form on the terminal and sent to the server. The output is the collected basic information and activity data.

[0484] The server generates a personalized care plan using a generative AI model based on the preprocessed health data and emotion data. It receives the preprocessed data and a specific model as input. Specifically, it generates a care plan that includes recommended activities such as "walk 7,000 steps per day" and "communicate with friends three times a week." The output is a personalized care plan.

[0485] Step 6: View Care Plan

[0486] The device displays the generated personalized care plan on the home screen or a dedicated application screen. It receives the generated care plan as input. Specifically, it visualizes the activities the user should perform that day and recommended health behaviors. The output is a visually presented care plan for the user.

[0487] Step 7: Anticipate health needs

[0488] The server analyzes current health, lifestyle, and emotional data and predicts future health risks using a generative AI model. It receives real-time data and a generative AI model as input. Specifically, it predicts the risk of nutritional deficiencies and high blood pressure. The output is health risk prediction data.

[0489] Based on the prediction results, the system suggests necessary health management actions to the user. It receives health risk prediction data as input. Specifically, it suggests taking specific supplements or undergoing regular medical checkups. The output is the suggested health management actions.

[0490] Step 8: Working with notifications

[0491] The terminal receives the prediction results and the personalized care plan from the server and sends them to the care staff's terminal. The terminal receives the prediction results and the care plan as input. Specifically, it displays a notification on the care staff's smartphone or tablet. The output is notification information for the care staff.

[0492] Based on the received information, care staff plan and implement specific care responses for the user. As input, they receive care plans and health risk prediction data. Specifically, they adjust the user's visit schedule and instruct specific health management actions. The output is the implemented care actions.

[0493] Step 9: Gather feedback

[0494] The user inputs the results of the care plan, their impressions, and changes in their physical condition via their device, and provides feedback to the server. Feedback data from the user is received as input. Specifically, feedback such as "I felt good walking today" is entered into a dedicated app. The output is the collected feedback data.

[0495] The server analyzes the collected feedback and uses it to improve and retrain the generative AI model. It receives the feedback data as input, adds new feedback, and retrains the model. The output is an improved generative AI model.

[0496] (Application example 2)

[0497] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart device 14 will be referred to as a "terminal."

[0498] Providing an environment where elderly people can travel safely and with peace of mind is an important issue in modern society. However, conventional self-driving vehicles have difficulty providing optimal travel plans that take into account the health and emotional state of each elderly person. Furthermore, systems that can respond quickly in emergencies are inadequate. Therefore, there is a need for a system that can balance health management and transportation safety for elderly people.

[0499] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[0500] In this invention, the server includes means for collecting user health data, means for preprocessing the collected health data, means for generating an individualized care plan using a generative AI model based on the preprocessed data, means for displaying the generated care plan, means for predicting future health needs using the generative AI model, means for notifying care staff of the prediction results and the care plan, means for collecting user feedback and using it to retrain the model, means for proposing an optimal travel plan based on the user's health and emotional state in the autonomous vehicle, and means for detecting abnormal health conditions and sending emergency notifications in real time. This allows for providing an optimal travel plan that takes into account the individual health and emotional states of the elderly person, enabling rapid response in emergencies.

[0501] "Health data" refers to a user's steps, heart rate, sleep data, medical records, health checkup results, and daily life log information.

[0502] "Preprocessing" refers to the process of removing noise from collected data, correcting missing values ​​and outliers, and extracting features.

[0503] A "generative AI model" is a model trained using deep learning algorithms to predict personalized care plans and health needs based on a user's features.

[0504] "Individualized care plan" refers to a plan that provides daily activity instructions and recommends healthy behaviors based on the user's health condition and lifestyle.

[0505] "Emotional state" refers to the type of emotion (joy, sadness, anger, etc.) assessed from the user's facial expression and tone of voice.

[0506] An "optimal travel plan" refers to a plan that suggests the most comfortable and safe travel route and travel conditions for the user based on the user's health and emotional state.

[0507] "Emergency notification" refers to a system that detects abnormalities in a user's health condition in real time and promptly notifies care staff and medical institutions.

[0508] "Feedback" refers to users reporting the results of implementing a care plan and changes in their physical condition, providing data to improve the accuracy of the generative AI model based on this.

[0509] An "autonomous vehicle" is a vehicle that uses artificial intelligence to drive autonomously and reach its destination without the user having to operate it.

[0510] A specific system for implementing this invention is an application running in a self-driving vehicle that collects health and emotional data from users and provides optimal travel plans for elderly people based on that data.

[0511] Hardware and software used

[0512] Hardware:

[0513] Autonomous vehicle control unit

[0514] Smartwatches and fitness trackers worn by seniors

[0515] Cameras and microphones installed inside the vehicle

[0516] Sensor device (vehicle vital signs monitor)

[0517] software:

[0518] Generative AI models (e.g., deep learning models using TensorFlow or PyTorch)

[0519] Natural language processing technology (e.g., NLTK, spaCy)

[0520] Emotion engines (e.g. emotion recognition APIs such as Face++ and Affectiva)

[0521] What the program does

[0522] server:

[0523] The server periodically collects step counts, heart rate, and sleep data from smartwatches and fitness trackers.

[0524] It connects with medical institution databases to obtain users' medical records and health checkup results.

[0525] Emotion data is obtained from an emotion engine that uses a camera and microphone to recognize the user's face and analyze their tone of voice.

[0526] Data preprocessing:

[0527] The server removes noise from the collected data and corrects missing and outlier values.

[0528] Standardize the data and extract daily features (e.g., average daily steps, resting heart rate, total calorie intake, emotional data).

[0529] Generative AI models:

[0530] The generative AI model generates personalized care plans based on pre-processed data, while a model for predicting health needs is also built at the same time.

[0531] Emotional data is reflected in this model, enabling highly accurate predictions.

[0532] User Interface:

[0533] The generated care plan and predicted health needs are displayed on a display inside the vehicle.

[0534] Through this display, users can check their activity and health management instructions for the day.

[0535] Emergency Response:

[0536] The server monitors the user's current health data in real time.

[0537] If an abnormality is detected, the emergency notification system will be activated quickly to notify care staff and medical institutions.

[0538] Specific examples

[0539] 1. Data Collection and Analysis

[0540] Before an elderly person gets into an autonomous vehicle, their smartwatch sends their recent health data (step count, heart rate, and sleep data) to a server.

[0541] Cameras and microphones inside the vehicle collect user emotional data, which is then analyzed by an emotion engine.

[0542] 2. Generating a travel plan

[0543] Based on the collected data, the generative AI model suggests optimal travel routes for users (for example, routes that pass through parks or routes with less congestion).

[0544] The trip plan is displayed on the vehicle's display, allowing the user to view the suggested route and any important points to note.

[0545] 3. Emergency Response

[0546] Health data is monitored in real time during the journey, and if any abnormalities are detected, the vehicle will automatically stop and notify care staff or medical institutions.

[0547] Prompt Sentence Examples

[0548] Prompt: The user is currently a senior citizen. Suggest the best travel route based on his recent health and emotional data. He has low step counts and is feeling depressed, so suggest a relaxing route through the park.

[0549] This invention is a system that enables elderly people to travel safely and comfortably according to their individual health and emotional state. Real-time data monitoring and rapid emergency response provide an environment where autonomous vehicles can be used with greater peace of mind.

[0550] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[0551] Step 1: Data collection

[0552] The server periodically collects step count, heart rate, and sleep data from the user's smartwatch or fitness tracker. It then connects with medical institution databases to obtain the user's medical records and health checkup results. It also uses the device's camera and microphone to analyze the user's facial recognition and tone of voice to assess their emotional state. This data is then input into the server, which outputs a comprehensive dataset on the user's health and emotional state based on the input data.

[0553] Step 2: Data Preprocessing

[0554] The server performs noise removal on the data collected in step 1 and corrects missing and outlier values. Next, it standardizes the data and extracts features such as average daily steps, resting heart rate, total calorie intake, and emotional data. It also converts text data (user feedback and medical records) into structured data using natural language processing technology. The standardized dataset is output.

[0555] Step 3: Generate an individualized care plan

[0556] The server generates a personalized care plan based on the pre-processed data using a generative AI model. The model is trained on a large dataset for elderly care. Based on the input data, the server outputs a care plan that includes daily activity instructions and recommended health behaviors.

[0557] Step 4: Reflecting emotional data

[0558] The server also incorporates emotional data into the generation of personalized care plans, providing highly accurate plans. Data obtained from the emotion engine is integrated with health data and analyzed using a generative AI model. A care plan corresponding to the emotional state based on the input data is output.

[0559] Step 5: View the care plan

[0560] The device then displays the generated care plan on the vehicle's internal display, where the user can view the day's activities and health management instructions. Based on the displayed data, the user can adjust their daily activities.

[0561] Step 6: Real-time health monitoring

[0562] As in step 1, the server monitors the user's health data in real time. If an abnormality is detected, an emergency notification system is activated immediately to notify care staff and medical institutions. This allows abnormality detection and response based on the input data.

[0563] Step 7: Optimize your travel plan

[0564] The server proposes an optimal travel plan based on the user's health and emotional state. It uses a generative AI model to calculate comfortable travel routes and travel conditions. Based on the input data, an optimized travel plan is output.

[0565] Step 8: Implement emergency response

[0566] The device detects abnormalities in the user's health condition in real time and sends an emergency notification immediately. The vehicle automatically stops and notifies pre-determined care staff and medical institutions. Emergency response is carried out based on the input data.

[0567] Step 9: Gather feedback

[0568] Users input the results of their care plan implementation and changes in their physical condition via their device. The server collects this feedback and uses it to retrain the generative AI model, improving the accuracy and effectiveness of the entire system. The retrained model is output based on the input data.

[0569] The specific processing unit 290 transmits the result of the specific processing to the smart device 14. In the smart device 14, the control unit 46A causes the output device 40 to output the result of the specific processing. The microphone 38B acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 38B to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[0570] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[0571] In the above embodiment, an example in which the specific process is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific process may be performed by the smart device 14.

[0572] [Second embodiment]

[0573] FIG. 3 shows an example of the configuration of a data processing system 210 according to the second embodiment.

[0574] 3, the data processing system 210 includes the data processing device 12 and smart glasses 214. An example of the data processing device 12 is a server.

[0575] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[0576] The smart glasses 214 include a computer 36, a microphone 238, a speaker 240, a camera 42, and a communication I / F 44. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, and the camera 42 are also connected to the bus 52.

[0577] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[0578] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[0579] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[0580] Fig. 4 shows an example of the main functions of the data processing device 12 and the smart glasses 214. As shown in Fig. 4, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[0581] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[0582] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[0583] In the smart glasses 214, the reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[0584] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the smart glasses 214 will be referred to as the "terminal."

[0585] MODE FOR CARRYING OUT THE INVENTION

[0586] The present invention is a system for creating personalized care plans and predicting health needs using generative AI in Japan's elderly care service management system. Specific embodiments of the system are described below.

[0587] System Configuration

[0588] This system consists of users (elderly people or their caregivers), a server, and terminals (smartphones, tablets, PCs, etc.).

[0589] Program processing flow

[0590] 1. Data Collection

[0591] server

[0592] The server periodically collects step counts, heart rate, and sleep data from smartwatches and fitness trackers worn by users.

[0593] It connects with medical institution databases to obtain users' medical records and health checkup results.

[0594] It also collects daily activity logs, such as dietary habits, exercise, and medication intake, which are entered by the user via the device.

[0595] 2. Data Preprocessing

[0596] server

[0597] Noise is removed from the collected data and missing values ​​are corrected.

[0598] The total number of steps taken each day is calculated based on the step count data from the smartwatch, and the resting heart rate is calculated based on the heart rate data.

[0599] Text data (e.g., meal details) is converted into structured data using natural language processing technology.

[0600] 3. Model Building

[0601] server

[0602] A generative AI model is trained using large amounts of pre-collected elderly care and medical data, and the model uses deep learning techniques.

[0603] The model generates personalized care plans and predicts future health needs based on the user's features.

[0604] 4. Generate personalized care plans

[0605] User

[0606] Users enter basic information (age, gender, medical history) and daily activity data through the device.

[0607] The server uses this information to generate an individualized care plan using a generative AI model.

[0608] The generated care plan is displayed on the device, allowing the user to check daily activity plans and recommended health behaviors.

[0609] 5. Anticipating health needs

[0610] server

[0611] The server uses a generative AI model to predict future health risks (e.g., risk of developing nutritional deficiencies or high blood pressure) based on the user's current health and lifestyle data.

[0612] The prediction results are communicated to the user and care staff in real time.

[0613] 6. Collaboration with care staff

[0614] Terminal

[0615] The individualized care plan and health needs prediction results received from the server are sent to the care staff's terminal.

[0616] The care staff adjusts specific care responses for the user based on the information received.

[0617] 7. Feedback Loops

[0618] User

[0619] The user provides feedback to the server via their device about the results of the care plan, their impressions, and changes in their physical condition.

[0620] The server analyzes the collected feedback and uses it to retrain the generative AI model.

[0621] Specific examples

[0622] Example 1: Data collection and preprocessing

[0623] The server automatically collects nighttime sleep data and daytime step count data from the smartwatch worn by the user every day, and cleanses this data to extract useful features.

[0624] Example 2: Generating personalized care plans

[0625] When a user enters their regular health check data into the terminal, the server uses that data to generate an individualized care plan, such as "walking 8,000 steps per day" or "light aerobic exercise twice a week," and displays it to the user.

[0626] Example 3: Predicting health needs

[0627] The server analyzes data entered by the user, such as a recent loss of appetite, and evaluates the estimated risk of nutritional deficiency. The generative AI model then suggests taking a multivitamin supplement and sends an alert to care staff.

[0628] The system allows users to implement personalized care plans to maintain and improve their health, and it also allows care staff to respond more accurately, improving the quality of elderly care.

[0629] The processing flow will be explained below.

[0630] Step 1: Data collection

[0631] server

[0632] The server works in conjunction with smartwatches and fitness trackers to periodically collect the user's steps, heart rate, and sleep data.

[0633] It connects to medical institution databases via API to obtain users' past medical records and regular health checkup results.

[0634] Users input their daily activities and dietary information into the device, and this data is also sent to the server.

[0635] Step 2: Data Preprocessing

[0636] server

[0637] The server cleanses the collected data, corrects missing and outlier values, and removes inappropriate data.

[0638] Standardize the data and extract features (e.g., average daily steps, resting heart rate, total calorie intake).

[0639] Text data (meal details, user feedback) is converted into structured data using natural language processing technology.

[0640] Step 3: Model Building

[0641] server

[0642] The server trains generative AI models using large datasets for elderly care.

[0643] Deep learning algorithms are used for training to build models that generate personalized care plans from user features and predict health needs.

[0644] Evaluate the accuracy of the model and tune the parameters as needed.

[0645] Step 4: Generate an individualized care plan

[0646] User

[0647] Users enter their basic information (age, gender, medical history) and daily activity data through the device.

[0648] server

[0649] The server collects the information submitted by the user and generates a personalized care plan using a generative AI model.

[0650] The generated care plan includes daily activity instructions and recommended health behaviors (e.g., aiming for 5,000 steps per day, eating a balanced diet).

[0651] Step 5: View Care Plan

[0652] Terminal

[0653] The designated care plan is displayed on the device screen, allowing the user to check the day's activities and health management instructions.

[0654] Step 6: Anticipate health needs

[0655] server

[0656] By analyzing a user's current health and lifestyle data, a generative AI model predicts future health risks (e.g., nutritional deficiencies, risk of chronic diseases).

[0657] Based on the prediction results, the system suggests necessary health management actions to the user (e.g., taking specific supplements, undergoing regular medical checkups).

[0658] Step 7: Working with notifications

[0659] Terminal

[0660] The prediction results and individualized care plan received from the server are sent to the care staff's terminal.

[0661] Care staff

[0662] Based on the information received, the care staff plans and implements specific care responses for the user.

[0663] Step 8: Gather feedback

[0664] User

[0665] Users input the results of the care plan, their impressions, and changes in their physical condition through their terminals, and provide feedback to the server.

[0666] server

[0667] The collected feedback is analyzed and used to improve and retrain the generative AI model.

[0668] Through feedback loops, the accuracy and effectiveness of the model and the system as a whole are continually improved.

[0669] Example 1

[0670] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0671] To improve the quality of elderly care, it is important to accurately generate individualized care plans and predict future health risks. However, conventional systems have difficulty centrally managing and effectively analyzing a user's diverse health data. Furthermore, they lack sufficient collaboration with care staff, making it difficult to provide appropriate care. Furthermore, they lack a mechanism for appropriately incorporating user feedback and updating the model, making it difficult to respond to ever-changing health conditions.

[0672] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[0673] In this invention, the server includes: means for collecting health data from a user's activity tracking device; means for acquiring medical data in cooperation with a medical institution's database; means for collecting daily activity data entered by the user; means for performing noise reduction and missing value correction on the collected health data; means for calculating the total number of steps per day based on step count data and the resting heart rate based on heart rate data; means for converting text data into structured data using natural language processing technology; means for training a generative AI model using elderly care data and medical data; means for generating an individualized care plan using the generative AI model; means for displaying the generated care plan; means for predicting future health needs using the generative AI model; means for notifying care staff of the prediction results and the care plan; means for collecting user feedback and using it to retrain the model; means for care staff to adjust care responses to the user based on the care plan and prediction results received; and means for the user to provide feedback. This enables the generation of individualized care plans and prediction of health risks, improving the quality of elderly care. Furthermore, by appropriately incorporating user feedback and updating the model, care can be provided that is always tailored to the latest health status.

[0674] "User" refers to the elderly person or their caregiver who uses the system.

[0675] An "activity measuring device" is a device for collecting a user's physical activity data, and includes smart watches, fitness trackers, etc.

[0676] "Health data" refers to information such as a user's number of steps, heart rate, sleep data, medical records, health checkup results, dietary habits, exercise amount, and medication intake.

[0677] "Server" refers to the computer system that collects, preprocesses, analyzes, and stores user data, and trains and runs generative AI models.

[0678] "Noise reduction" refers to the process of removing unnecessary or erroneous information from collected data.

[0679] "Missing value correction" refers to the process of making predictions or inferences to fill in missing values ​​in data.

[0680] "Natural language processing technology" refers to technology for converting text data into a format that is easy for machines to understand.

[0681] "Generative AI model" refers to an AI (artificial intelligence) model that uses large amounts of data to generate personalized care plans and predict future health needs.

[0682] "Individualized care plan" refers to a care plan created based on a user's specific health condition and lifestyle habits.

[0683] "Health needs prediction" refers to the process of predicting future health risks and care needs based on a user's current data.

[0684] "Care staff" refers to professional people who provide care for older people.

[0685] "Feedback" refers to information provided by the user regarding the results of implementing the care plan, their impressions, and changes in their physical condition.

[0686] "Model retraining" refers to the process of updating a generative AI model with newly collected data and feedback to improve its accuracy.

[0687] This invention is a system for creating personalized care plans and predicting health needs using a generative AI model in an elderly care service management system in Japan. The system consists of users (elderly people or their caregivers), a server, and devices (smartphones, tablets, personal computers, etc.).

[0688] System Configuration

[0689] The system is configured as follows:

[0690] server

[0691] The server is responsible for:

[0692] 1. Periodically collect step count, heart rate, and sleep data from the user's activity tracking device (smartwatch or fitness tracker).

[0693] 2. Link with medical institution databases to obtain users' medical records and health checkup results.

[0694] 3. It also collects daily activity logs, such as dietary details, exercise volume, and medication intake, entered by the user through the device.

[0695] 4. Remove noise from the collected data and correct missing values.

[0696] 5. Calculate the total number of steps taken each day based on the step count data from the smartwatch, and calculate the resting heart rate based on the heart rate data.

[0697] 6. Convert text data (e.g., meal details) into structured data using natural language processing technology.

[0698] 7. Train generative AI models using elderly care and medical data, which use deep learning techniques.

[0699] 8. The model generates a personalized care plan and predicts future health needs based on the user's features.

[0700] Terminal

[0701] The device is used by users and care staff and has the following functions:

[0702] 1. The user enters basic information (age, gender, medical history) and daily activity data through the device.

[0703] 2. The device displays the personalized care plan generated using the generative AI model.

[0704] 3. Notify users and care staff in real time of health risk prediction results and the implementation results of care plans.

[0705] 4. Care staff will coordinate specific care responses for the user based on the information received.

[0706] User

[0707] Users provide daily activity data and feedback to the system:

[0708] 1. Users send health data to a server via their smartwatch or fitness tracker.

[0709] 2. The user enters information such as dietary habits and medication intake into the terminal application.

[0710] 3. Provide feedback on the results of the care plan, impressions, and changes in physical condition via the device application.

[0711] Specific examples

[0712] Example 1: Data collection and preprocessing

[0713] The server automatically collects nighttime sleep data and daytime step count data from the smartwatch worn by the user every day, and cleanses this data to extract useful features.

[0714] Example 2: Generating personalized care plans

[0715] When a user enters their regular health check data into the terminal, the server uses that data to generate an individualized care plan, such as "walking 8,000 steps per day" or "light aerobic exercise twice a week," and displays it to the user.

[0716] Example 3: Predicting health needs

[0717] The server analyzes data entered by the user, such as a recent loss of appetite, and assesses the estimated risk of nutritional deficiency. The generative AI model then suggests taking a multivitamin supplement and sends an alert to care staff. This system allows users to implement personalized care plans to maintain and improve their health. It also allows care staff to respond more accurately, improving the quality of elderly care.

[0718] The flow of the identification process in the first embodiment will be described with reference to FIG.

[0719] Step 1: Data collection

[0720] Server collects smartwatch data

[0721] Input: Step count, heart rate, and sleep data from the user's smartwatch

[0722] Data processing: Collect data from the smartwatch periodically and store it in a database

[0723] Output: Raw health data

[0724] Specific operation: When the user syncs the smartwatch every morning, the server receives the night's sleep data and the previous day's step count data and stores them in a database.

[0725] Acquisition of medical data by the server

[0726] Input: Medical records and health check results from medical institution databases

[0727] Data processing: Data is acquired through the medical institution's API and saved as user health information.

[0728] Output: Retrieved medical data

[0729] Specific operation: Outside of the medical institution's scheduled maintenance period, the server sends an API request to obtain the user's latest blood test results and diagnostic data.

[0730] Manual data collection by the server

[0731] Input: Dietary information, exercise amount, and medication intake status entered by the user via the device

[0732] Data processing: Collect input data in real time and save it in a database

[0733] Output: Daily activity data

[0734] Specific operation: After dinner, the user uses the terminal application to input the details of their meal and medication intake, and the data is sent to the server in real time.

[0735] Step 2: Data Preprocessing

[0736] Server-based noise removal and missing value correction

[0737] Input: Collected health data

[0738] Data processing: data denoising and missing value imputation

[0739] Output: Cleaned health data

[0740] Specific operation: The server cleanses the data obtained from the smartwatch and fills in missing daily step count data with data from the previous day.

[0741] Server recalculation and structuring of data

[0742] Input: Data after noise removal and correction

[0743] Data processing: Counting step count data, calculating resting heart rate, structuring text data

[0744] Output: Recalculated and structured data

[0745] Specific operation: The server aggregates the step count data and calculates the total number of steps taken for the day. It also analyzes the text of meal descriptions entered by the user, such as "salad and grilled chicken," and extracts information about calories and nutrients as structured data.

[0746] Step 3: Model Building

[0747] Server-based training of generative AI models

[0748] Input: Elderly care data, medical data

[0749] Data processing: Using deep learning techniques to train AI models

[0750] Output: A trained generative AI model

[0751] How it works: Generative AI models are trained using historical medical datasets and iteratively trained until error is minimized.

[0752] Step 4: Generate an individualized care plan

[0753] User enters basic information

[0754] Input: User's basic information (age, gender, medical history), daily activity data

[0755] Data processing: Enter basic information into the system and send it to the server

[0756] Output: User basic information data

[0757] Specific operation: A user logs in to the application and enters their age, gender, and medical history on the profile screen.

[0758] Server-based care plan generation

[0759] Input: User's basic information and daily activity data

[0760] Data processing: Generative AI models are used to generate personalized care plans

[0761] Output: Individualized Care Plan

[0762] Specific operation: The server calls up an AI model based on the basic information entered and generates a daily activity schedule, recommended exercises, and a meal plan.

[0763] Displaying care plans on devices

[0764] Input: Generated personalized care plan

[0765] Data processing: Push notification of care plan to device

[0766] Output: The care plan displayed to the user

[0767] What it does: The care plan is pushed to the device, and when the user opens the app, a detailed activity plan is displayed on the dashboard.

[0768] Step 5: Anticipate health needs

[0769] Server-based health risk prediction

[0770] Input: User's current health and lifestyle data

[0771] Data processing: Predicting future health risks using generative AI models

[0772] Output: Health risk prediction results

[0773] How it works: Analyzes recent dietary data to assess the user's risk of vitamin deficiency. If the deficiency persists, the AI ​​model predicts the risk and generates an alert.

[0774] Real-time notifications

[0775] Input: Health risk prediction results

[0776] Data processing: Real-time notification of prediction results

[0777] Output: Notification to user and care staff

[0778] Specific operation: A push notification is sent to the care staff's device, and an alert is displayed stating, "User A is at risk of high blood pressure."

[0779] Step 6: Collaborate with care staff

[0780] Sending care information by device

[0781] Input: Personalized care plan and health risk prediction results received from the server

[0782] Data processing: Sending information to care staff's terminal

[0783] Output: Care plan and health risk prediction information

[0784] Specific operation: Information from the server is sent to an app dedicated to care staff, who can check it in real time.

[0785] Care staff coordination

[0786] Input: Care plan and health risk prediction information

[0787] Data processing: Coordination of care responses

[0788] Output: Adjusted care plan

[0789] Specific operation: During regular visits, care staff use information obtained from the server to fine-tune the user's care plan.

[0790] Step 7: Feedback Loop

[0791] User feedback

[0792] Input: Care plan implementation results, impressions, changes in physical condition

[0793] Data processing: Input the feedback into the system and send it to the server

[0794] Output: User feedback information

[0795] Specific operation: When the user enters feedback into the app, such as "I've been walking for a while, and as a result, I feel better," that information is sent to the server.

[0796] Server data reuse

[0797] Input: Collected feedback data

[0798] Data processing: Retraining generative AI models

[0799] Output: Improved generative AI model

[0800] Specific operation: The server analyzes the collected feedback data and performs re-training to improve the accuracy of the AI ​​model.

[0801] (Application example 1)

[0802] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0803] In conventional elderly care services, collecting health data and creating care plans is often done manually, making it difficult to provide personalized care. Furthermore, there is a lack of means to monitor health conditions in real time and quickly respond as needed. This poses a particular challenge when an elderly person becomes ill while traveling or when emergency care is required.

[0804] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[0805] In this invention, the server includes means for collecting user health data, means for preprocessing the collected health data, means for generating an individualized care plan based on the preprocessed data using a generative AI model, means for displaying the generated care plan in the vehicle, means for predicting future health needs using the generative AI model, means for notifying a caregiver of the prediction results and the care plan, means for collecting feedback from the user and using it to retrain the model, means for optimizing a driving route according to the user's health condition, and means for detecting and notifying abnormal health data in real time, thereby enabling the provision of an individualized care plan, real-time health condition monitoring, and appropriate responses even while traveling.

[0806] "User" refers to an individual about whom health data is collected, particularly the elderly and those in need of health management.

[0807] "Health data" refers to information that indicates an individual's health status, such as heart rate, number of steps, sleep data, dietary content, amount of exercise, and medication intake.

[0808] "Preprocessing" refers to the processing of collected health data, such as noise removal, missing value correction, and data structuring.

[0809] A "generative AI model" is an artificial intelligence model trained using deep learning techniques to generate personalized care plans based on collected health data and predict future health needs.

[0810] A "personalized care plan" is a specific action plan created based on the user's individual health data and lifestyle, with the aim of maintaining or improving the health of a specific user.

[0811] An "autonomous vehicle" is a vehicle that is capable of driving independently and is equipped with the functionality to collect and process health data.

[0812] "Display means" refers to a device or function for visually showing the generated care plan to the user or caregiver (e.g., an in-car display or a smartphone).

[0813] "Health needs" are the activities and care requirements necessary to maintain or improve the user's future health status.

[0814] "Care aides" are professionals and caregivers who provide care to older people and people with health care needs.

[0815] "Notification means" refers to a device or system that notifies caregivers of the generated care plan and the predicted health needs (e.g., a smartphone app or an in-car system).

[0816] "Feedback" is information that the user reports, such as the results of implementing the care plan, their impressions, and changes in their physical condition.

[0817] "Model retraining" is a machine learning process that uses collected feedback data to further improve a generative AI model.

[0818] "Route optimization" is the process of selecting the optimal driving route and rest points based on the user's health condition.

[0819] "Real-time detection" is a function that can immediately recognize and respond to abnormalities in health data when they occur.

[0820] The present invention relates to a system for providing personalized health care in autonomous vehicles, particularly designed for the elderly and people with health care needs. Specific embodiments of the system are described below.

[0821] System Configuration

[0822] This system consists of a user (such as an elderly person), a server (the autonomous vehicle's infotainment system), a device (such as a smartwatch or fitness tracker) for collecting health data, and a caregiver (such as a caregiver or medical staff).

[0823] Program processing flow

[0824] Data collection

[0825] The server collects real-time health data such as heart rate, step count, and sleep data from the smartwatch or fitness tracker worn by the user. It also connects to medical institution databases to obtain the user's medical records and health checkup results. It also collects dietary information and daily activity logs entered by the user through a terminal inside the autonomous vehicle.

[0826] Data Preprocessing

[0827] The server removes noise from the collected health data and corrects missing values. It calculates the total number of steps taken each day based on the step count data from the smartwatch and calculates the resting heart rate based on the heart rate data. Text data such as meal details is converted into structured data using natural language processing technology.

[0828] Model Building

[0829] The server uses large amounts of pre-collected elderly care and medical data to train a generative AI model, which uses deep learning techniques to generate personalized care plans and predict future health needs based on the user's features.

[0830] Personalized care plan generation

[0831] Users enter basic information (age, gender, medical history) and daily activity data through the infotainment system. The server uses this information to generate a personalized care plan using a generative AI model. The generated care plan is displayed on a display inside the autonomous vehicle, allowing users to check their daily activity plan and recommended health behaviors.

[0832] Anticipating health needs

[0833] The server uses a generative AI model to predict future health risks (e.g., risk of developing nutritional deficiencies or high blood pressure) based on the user's current health and lifestyle data. The prediction results are notified to the user and caregivers in real time.

[0834] Cooperation with care staff

[0835] The personalized care plan and health needs prediction results received from the server are sent to the caregiver's device, who then adjusts specific care responses for the user based on the received information.

[0836] Feedback Loop

[0837] Users provide feedback to the server via their devices about the results of implementing the care plan, their impressions, and changes in their physical condition. The server analyzes the collected feedback and uses it to retrain the generative AI model.

[0838] Hardware and software used

[0839] Smartwatch: Collects heart rate, steps, and sleep data.

[0840] Autonomous vehicles: Infotainment systems for collecting and processing health data.

[0841] Server: Preprocessing data, training and running generative AI models.

[0842] Deep learning framework (TensorFlow or PyTorch): Implementing and training generative AI models.

[0843] Natural language processing technologies (SpaCy, NLTK, etc.): Structuring text data.

[0844] Specific examples

[0845] For example, if a user's heart rate increases while riding in an autonomous vehicle, the autonomous vehicle's system will collect heart rate data in real time and provide a care plan to "take a deep breath and relax" and a notification recommending a break at the next service area.

[0846] Prompt Sentence Examples

[0847] "The system monitors passengers' health status in real time based on heart rate and step count data during the ride, and immediately proposes an appropriate care plan if an abnormality is detected. It also suggests optimal routes and rest points depending on the passenger's condition."

[0848] The system enables personalized care plans, real-time health monitoring and appropriate responses, even while on the move.

[0849] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[0850] Step 1:

[0851] A user wears a smartwatch or fitness tracker and gets into an autonomous vehicle. The device collects real-time heart rate, step count, and sleep data from the smartwatch and sends them to a server. The input of this step is the health data from the smartwatch, and the output is the raw data sent to the server.

[0852] Step 2:

[0853] The server preprocesses the received health data. Specifically, it removes noise and corrects missing values. For example, it complements missing data based on past data and removes obviously erroneous data. The input is the raw data collected in step 1, and the output is the cleansed health data.

[0854] Step 3:

[0855] The server uses a generative AI model to generate a personalized care plan based on the preprocessed data. For example, it may recommend deep breathing if the heart rate is high or light exercise if the step count is low. The input for this step is the preprocessed health data, and the output is a personalized care plan.

[0856] Step 4:

[0857] The generated care plan is presented to the user through the autonomous vehicle's infotainment system, where the user can visually review it and follow the instructions. The input is the generated personalized care plan, and the output is the care plan displayed to the user.

[0858] Step 5:

[0859] The server uses the generative AI model to predict the user's future health needs. For example, it evaluates the risk of rising blood pressure several months from now based on current data. The input of this step is the preprocessed health data, and the output is the predicted health risks and future health needs.

[0860] Step 6:

[0861] The server notifies the caregiver of the prediction results and the care plan. The caregiver can then adjust specific care responses for the user based on this information. The input is the prediction results and the care plan, and the output is a notification sent to the caregiver.

[0862] Step 7:

[0863] The server provides an interface for collecting feedback from users. Users input the results of implementing the care plan, their impressions, and changes in their physical condition. The input is the user's feedback, and the output is the feedback data stored on the server.

[0864] Step 8:

[0865] The server analyzes the collected feedback and uses it to retrain the generative AI model, which enables the model to provide more accurate care plans. The input is the feedback data, and the output is the retrained generative AI model.

[0866] Step 9:

[0867] The server optimizes the driving route according to the user's health condition. For example, it suggests appropriate rest areas according to the user's fatigue level. The input is health data detected in real time, and the output is the optimized driving route.

[0868] Step 10:

[0869] The server detects abnormal health data in real time and immediately notifies the user and caregivers, for example, sending an alert if the heart rate is abnormally high. The input is the health data collected in real time, and the output is the alert notification sent.

[0870] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[0871] MODE FOR CARRYING OUT THE INVENTION

[0872] The present invention is a system for creating personalized care plans and predicting health needs using generative AI and an emotion engine in an elderly care service management system in Japan. Specific embodiments of the system are described below.

[0873] System Configuration

[0874] This system consists of users (elderly people or their caregivers), a server, terminals (smartphones, tablets, PCs, etc.), and an emotion engine.

[0875] Program processing flow

[0876] 1. Data Collection

[0877] server

[0878] The server periodically collects step counts, heart rate, and sleep data from smartwatches and fitness trackers worn by users.

[0879] It connects with medical institution databases to obtain users' medical records and health checkup results.

[0880] It also collects daily activity logs, such as dietary habits, exercise, and medication intake, which are entered by the user via the device.

[0881] 2. Collecting Emotional Data

[0882] Emotion Engine

[0883] The emotion engine uses facial recognition technology to analyze the user's facial expressions and assess their emotional state (e.g., happy, sad, or angry).

[0884] Analyze the tone of the user's voice to recognize their emotional state.

[0885] 3. Data Preprocessing

[0886] server

[0887] Noise is removed from the collected health and emotion data, missing and outliers are corrected, and inappropriate data is removed.

[0888] The data is standardized and features (e.g., average daily steps, resting heart rate, total calorie intake, and emotional data) are extracted.

[0889] Text data (meal details, user feedback) is converted into structured data using natural language processing technology.

[0890] 4. Model Building

[0891] server

[0892] A generative AI model is trained using a deep learning algorithm on a large dataset for elderly care.

[0893] The model builds a model that generates an individualized care plan based on the user's features and a model that predicts health needs.

[0894] It also takes into account sentiment data, evaluates the accuracy of the model, and tunes parameters as needed.

[0895] 5. Generate personalized care plans

[0896] User

[0897] Users enter basic information (age, gender, medical history) and daily activity data through the device.

[0898] server

[0899] The server collects the emotion data from the emotion engine and the information submitted by the user, and generates a personalized care plan using a generative AI model.

[0900] The generated care plan includes daily activity instructions and recommended health behaviors (e.g., aiming for 5,000 steps per day, eating a balanced diet).

[0901] 6. Care plan display

[0902] Terminal

[0903] The care plan is displayed on the device screen, allowing users to check their activities and health management instructions for the day.

[0904] 7. Anticipating health needs

[0905] server

[0906] By analyzing a user's current health, lifestyle, and emotional data, a generative AI model predicts future health risks (e.g., risk of nutritional deficiencies and high blood pressure).

[0907] Based on the prediction results, the system suggests necessary health management actions to the user (e.g., taking specific supplements, undergoing regular medical checkups).

[0908] 8. Notifications and Collaboration

[0909] Terminal

[0910] The prediction results and individualized care plan received from the server are sent to the care staff's terminal.

[0911] Care staff

[0912] Based on the information received, the care staff plans and implements specific care responses for the user.

[0913] 9. Feedback Collection

[0914] User

[0915] Users input the results of the care plan, their impressions, and changes in their physical condition through their terminals, and provide feedback to the server.

[0916] server

[0917] The collected feedback is analyzed and used to improve and retrain the generative AI model.

[0918] Through feedback loops, the accuracy and effectiveness of the model and the system as a whole are continually improved.

[0919] Specific examples

[0920] Example 1: Data collection and preprocessing

[0921] The user collects daily step counts and heart rate data from the smartwatch, and the server cleans the data to extract useful features. At the same time, the emotion engine collects the user's emotional data using the device's camera and microphone.

[0922] Example 2: Generating personalized care plans

[0923] Users who tend to be isolated input data from regular health checkups and daily life into the device, and the server generates an individualized care plan based on that data and emotional data obtained from the emotion engine, including recommendations such as "walking 7,000 steps a day" and "communicating with friends three times a week," and displays it to the user.

[0924] Example 3: Predicting health needs

[0925] The server analyzes data showing that the user enjoys walking but has recently been feeling depressed, and the generative AI model predicts that the user has a tendency toward seasonal depression. Reflecting the data from the emotion engine, the server recommends the user engage in daily light exercise and sun exposure, and notifies care staff to contact the user regularly.

[0926] The system allows users to receive scientifically-backed, personalized care plans and responses tailored to their emotional state, helping them maintain and improve their physical and mental health, while providing caregivers with more accurate information to provide specific care.

[0927] The processing flow will be explained below.

[0928] Step 1: Data collection

[0929] server

[0930] The server works in conjunction with smartwatches and fitness trackers to periodically collect the user's steps, heart rate, and sleep data.

[0931] It connects to medical institution databases via API to obtain users' past medical records and regular health checkup results.

[0932] Daily activity logs, such as what the user eats, how much exercise they do, and what medications they take, are also sent to the server.

[0933] Step 2: Collecting emotion data

[0934] Emotion Engine

[0935] The emotion engine uses the device's camera to capture the user's facial expressions and analyzes their emotional state (happiness, sadness, anger, etc.) using facial recognition technology.

[0936] When a user speaks using the device, the tone of their voice is analyzed through a microphone to recognize their emotional state.

[0937] Step 3: Data Preprocessing

[0938] server

[0939] The server cleanses the collected health and emotion data and corrects missing or outlier values.

[0940] The data is standardized and features (e.g., average daily steps, resting heart rate, total calorie intake, and emotional data) are extracted.

[0941] Text data (meal details, user feedback) is converted into structured data using natural language processing technology.

[0942] Step 4: Model Building

[0943] server

[0944] A generative AI model is trained using a large dataset for elderly care, using deep learning algorithms.

[0945] The model will be constructed to generate an individualized care plan based on the user's features, and to predict health needs.

[0946] The accuracy of the model is evaluated taking into account sentiment data and parameters are adjusted as needed.

[0947] Step 5: Generate an individualized care plan

[0948] User

[0949] Users enter their basic information (age, gender, medical history) and daily activity data into the device.

[0950] server

[0951] The server collects the emotion data from the emotion engine and the information submitted by the user, and generates a personalized care plan using a generative AI model.

[0952] The user's plan includes daily activity instructions and recommended health behaviors (e.g., aiming for 5,000 steps per day, adjusting diet, and taking mental health measures).

[0953] Step 6: View Care Plan

[0954] Terminal

[0955] The generated care plan is displayed on the terminal screen, and the user can check the day's activities and health management instructions.

[0956] Step 7: Anticipate health needs

[0957] server

[0958] The server analyzes the user's current health, lifestyle, and emotional data, and uses a generative AI model to predict future health risks (e.g., risk of nutritional deficiencies and high blood pressure).

[0959] Based on the prediction results, the system suggests necessary health management actions to the user (e.g., taking specific supplements, undergoing regular medical checkups).

[0960] Step 8: Working with notifications

[0961] Terminal

[0962] The prediction results and individualized care plan received from the server are sent to the care staff's terminal.

[0963] Care staff

[0964] Based on the information received, the care staff plans and implements specific care responses for the user.

[0965] Step 9: Gather feedback

[0966] User

[0967] The user inputs the results of the care plan, their impressions, and any changes in their physical condition into the terminal and provides feedback to the server.

[0968] server

[0969] The collected feedback is analyzed and used to improve and retrain the generative AI model.

[0970] Through feedback loops, the accuracy and effectiveness of the model and the system as a whole are continually improved.

[0971] Example 2

[0972] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[0973] In elderly care services, it is challenging to provide care plans that take into account the user's individual health and emotional state. While conventional systems focus on collecting and analyzing health data, incorporating the user's emotional data is needed to provide more accurate personalized care plans. However, such systems face challenges due to the complexity of data collection and the need to assess the user's emotional state in real time. Furthermore, a means of clearly communicating the generated care plan and health risk predictions to the user is also required, and a mechanism for seamlessly aggregating subsequent feedback and retraining the model is lacking.

[0974] The identification processing by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting user health data, means for preprocessing the collected health data, means for generating an individualized care plan using a generative AI model based on the preprocessed data, means for collecting user emotion data using facial recognition technology and voice analysis technology, means for preprocessing the emotion data, means for generating an individualized care plan using the preprocessed emotion data and health data, means for displaying the generated care plan, means for predicting future health needs using the generative AI model, means for notifying care staff of the prediction results and the care plan, and means for collecting feedback from the user and using it to retrain the model. This makes it possible to provide an accurate individualized care plan based on the user's health condition and emotional state, and is expected to improve the accuracy of health risk predictions and the quality of care.

[0975] A "user" is an entity that utilizes the system to provide health and emotional data and receive a personalized care plan.

[0976] "Health data" refers to data collected by users in their daily lives, such as the number of steps taken, heart rate, sleep data, dietary content, amount of exercise, and medication intake.

[0977] "Emotional data" refers to data relating to the emotional state of a user obtained from facial expressions and tone of voice collected using face recognition technology and voice analysis technology.

[0978] "Facial recognition technology" is a technology that uses a camera to analyze a user's facial expressions and assess their emotional state.

[0979] "Voice analysis technology" is a technology that uses a microphone to analyze the tone and intonation of a user's voice to assess their emotional state.

[0980] A "generative AI model" is an algorithmic model that uses collected health and emotional data to generate personalized care plans and predict future health needs.

[0981] A "personalized care plan" is a plan that includes health care instructions and activity recommendations that are optimal for a user, generated based on the user's health and emotional data.

[0982] "Preprocessing" refers to processing of collected data, such as noise removal, missing value correction, outlier removal, data standardization, and feature extraction.

[0983] "Feedback" is information provided by the user regarding the results of implementing the care plan, their impressions, and changes in their physical condition, and is used to retrain the model.

[0984] "Care staff" are people who plan and implement specific care responses for users based on the individualized care plans and health needs prediction results generated using the system.

[0985] "Notification" refers to the act of notifying the user or care staff of the generated care plan or prediction results via a terminal.

[0986] MODE FOR CARRYING OUT THE INVENTION

[0987] This invention is a system for creating personalized care plans and predicting health needs in elderly care services using generative AI and an emotion engine. This system consists of a user, a server, a terminal, and an emotion engine.

[0988] 1. Users

[0989] Users wear a smartwatch or fitness tracker to record their daily activity data. They also input their daily activity log, including their diet, exercise, and medication intake, via a device (smartphone, tablet, PC, etc.). Furthermore, the device's camera and microphone are used to record data on their emotional state.

[0990] 2. Server

[0991] The server does the following:

[0992] Data collection: The server periodically collects data from the user's smartwatch or fitness tracker and connects with medical institution databases to obtain medical records and health checkup results.

[0993] Emotional data collection: We collect user emotional data using facial recognition and voice analysis technologies.

[0994] Data preprocessing: Noise removal, missing value correction, outlier removal, standardization, feature extraction, and text data structuring are performed on the collected health and emotion data.

[0995] Model Building: Train generative AI models using large-scale elderly care datasets to build and evaluate models that generate personalized care plans and predict health needs.

[0996] Prediction and generation: Using a generative AI model, future health risks are predicted based on the user's features and emotional data, and an individualized care plan is generated.

[0997] Notification and collaboration: Prediction results and care plans are notified to care staff devices.

[0998] 3. Terminal

[0999] The terminal has the following features:

[1000] Feedback collection: Users input the results of the care plan, their impressions, and changes in their physical condition via their terminal and provide feedback to the server.

[1001] Care plan display: The generated care plan is displayed on the terminal screen, allowing the user to check the health care instructions for the day.

[1002] Notification: Display the prediction results and care plan sent from the server and notify the care staff.

[1003] Specific examples

[1004] Example 1: Data collection and preprocessing

[1005] The user records their daily steps and heart rate on their smartwatch, and the server cleans and extracts features from the data. The emotion engine collects the user's emotion data using the device's camera and microphone.

[1006] Example 2: Generating personalized care plans

[1007] Users who tend to be isolated input data from regular health checkups and daily life into the device, and the server generates an individualized care plan based on that data and emotional data obtained from the emotion engine, including recommendations such as "walking 7,000 steps a day" and "communicating with friends three times a week," and displays it to the user.

[1008] Example 3: Predicting health needs

[1009] The server analyzes data showing that the user enjoys walking but has recently been feeling depressed, and the generative AI model predicts that the user has a tendency toward seasonal depression. Reflecting the data from the emotion engine, the server recommends the user engage in daily light exercise and sun exposure, and notifies care staff to contact the user regularly.

[1010] Prompt Sentence Examples

[1011] "The user is 70 years old, their primary daily activity is walking, and the emotion engine has detected their recent emotional state as 'sad'. Please generate a personalized care plan for this user."

[1012] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1013] System program processing flow

[1014] Step 1: Data collection

[1015] The server periodically collects data such as step counts, heart rate, and sleep data from the smartwatch or fitness tracker worn by the user. As input, it receives biometric data from the smartwatch or fitness tracker. Specifically, it makes API calls from the smart device and stores the data in the server's database.

[1016] The server connects to the medical institution's database to retrieve the user's medical records and health checkup results. It receives data obtained from the medical institution's API as input. Specifically, it calls the API on a regular schedule, retrieves the necessary data, and stores it in the server's database.

[1017] Users input their daily activity logs, including their diet, exercise, and medication intake, via a device. The system receives inputs from users' manual entries and log data written on a daily basis. Specifically, data is entered through a dedicated application on the device and sent to a server. The output is the collected health data and daily activity log.

[1018] Step 2: Collecting emotion data

[1019] The emotion engine uses facial recognition technology to analyze video data acquired from the device's camera and analyze the user's facial expressions. It receives the video data acquired from the camera as input. Specifically, it uses a facial recognition algorithm to analyze the facial expressions and evaluate the user's emotional state, such as "happiness" or "sadness." The output is data indicating the emotional state.

[1020] The emotion engine analyzes audio data recorded using a microphone and recognizes emotions based on the user's tone of voice. It receives audio data from the microphone as input, analyzes the audio data using a voice analysis algorithm, and evaluates the user's emotional state. The output is audio data that indicates the user's emotional state.

[1021] Step 3: Data Preprocessing

[1022] The server performs noise removal on the collected health and emotion data. As input, it receives the collected raw data. Specifically, it detects outliers and missing values ​​and removes or corrects them. The output is the cleansed data.

[1023] The server normalizes the data and extracts features such as average daily steps, resting heart rate, total calorie intake, and emotional data. It receives the cleansed data as input and applies data normalization and feature extraction algorithms to convert it into analyzable data. The output is feature data for analysis.

[1024] The server converts text data (meal details, user feedback) into structured data using natural language processing technology. It receives text data entered by the user as input. Specifically, it applies a text analysis algorithm to convert the data into structured data, such as meal items. The output is structured text data.

[1025] Step 4: Model Building

[1026] The server trains a generative AI model using a large-scale dataset for elderly care. It receives a large amount of training data as input. Specifically, it uses a deep learning algorithm (e.g., TensorFlow or PyTorch) to train the model. The output is a trained generative AI model.

[1027] The server builds a model that generates an individualized care plan and a model that predicts health needs based on the user's features. It receives feature data as input. Specifically, it uses the generative AI model to create an optimal care plan and predict health risks for each user. The output is an individualized care plan generation model and a health needs prediction model.

[1028] Step 5: Generate an individualized care plan

[1029] The user inputs basic information (age, gender, medical history) and daily activity data through the terminal. The system receives basic information and daily activity data from the user as input. Specifically, the data is entered through an input form on the terminal and sent to the server. The output is the collected basic information and activity data.

[1030] The server generates a personalized care plan using a generative AI model based on the preprocessed health data and emotion data. It receives the preprocessed data and a specific model as input. Specifically, it generates a care plan that includes recommended activities such as "walk 7,000 steps per day" and "communicate with friends three times a week." The output is a personalized care plan.

[1031] Step 6: View Care Plan

[1032] The device displays the generated personalized care plan on the home screen or a dedicated application screen. It receives the generated care plan as input. Specifically, it visualizes the activities the user should perform that day and recommended health behaviors. The output is a visually presented care plan for the user.

[1033] Step 7: Anticipate health needs

[1034] The server analyzes current health, lifestyle, and emotional data and predicts future health risks using a generative AI model. It receives real-time data and a generative AI model as input. Specifically, it predicts the risk of nutritional deficiencies and high blood pressure. The output is health risk prediction data.

[1035] Based on the prediction results, the system suggests necessary health management actions to the user. It receives health risk prediction data as input. Specifically, it suggests taking specific supplements or undergoing regular medical checkups. The output is the suggested health management actions.

[1036] Step 8: Working with notifications

[1037] The terminal receives the prediction results and the personalized care plan from the server and sends them to the care staff's terminal. The terminal receives the prediction results and the care plan as input. Specifically, it displays a notification on the care staff's smartphone or tablet. The output is notification information for the care staff.

[1038] Based on the received information, care staff plan and implement specific care responses for the user. As input, they receive care plans and health risk prediction data. Specifically, they adjust the user's visit schedule and instruct specific health management actions. The output is the implemented care actions.

[1039] Step 9: Gather feedback

[1040] The user inputs the results of the care plan, their impressions, and changes in their physical condition via their device, and provides feedback to the server. Feedback data from the user is received as input. Specifically, feedback such as "I felt good walking today" is entered into a dedicated app. The output is the collected feedback data.

[1041] The server analyzes the collected feedback and uses it to improve and retrain the generative AI model. It receives the feedback data as input, adds new feedback, and retrains the model. The output is an improved generative AI model.

[1042] (Application example 2)

[1043] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the smart glasses 214 will be referred to as a "terminal."

[1044] Providing an environment where elderly people can travel safely and with peace of mind is an important issue in modern society. However, conventional self-driving vehicles have difficulty providing optimal travel plans that take into account the health and emotional state of each elderly person. Furthermore, systems that can respond quickly in emergencies are inadequate. Therefore, there is a need for a system that can balance health management and transportation safety for elderly people.

[1045] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1046] In this invention, the server includes means for collecting user health data, means for preprocessing the collected health data, means for generating an individualized care plan using a generative AI model based on the preprocessed data, means for displaying the generated care plan, means for predicting future health needs using the generative AI model, means for notifying care staff of the prediction results and the care plan, means for collecting user feedback and using it to retrain the model, means for proposing an optimal travel plan based on the user's health and emotional state in the autonomous vehicle, and means for detecting abnormal health conditions and sending emergency notifications in real time. This allows for providing an optimal travel plan that takes into account the individual health and emotional states of the elderly person, enabling rapid response in emergencies.

[1047] "Health data" refers to a user's steps, heart rate, sleep data, medical records, health checkup results, and daily life log information.

[1048] "Preprocessing" refers to the process of removing noise from collected data, correcting missing values ​​and outliers, and extracting features.

[1049] A "generative AI model" is a model trained using deep learning algorithms to predict personalized care plans and health needs based on a user's features.

[1050] "Individualized care plan" refers to a plan that provides daily activity instructions and recommends healthy behaviors based on the user's health condition and lifestyle.

[1051] "Emotional state" refers to the type of emotion (joy, sadness, anger, etc.) assessed from the user's facial expression and tone of voice.

[1052] An "optimal travel plan" refers to a plan that suggests the most comfortable and safe travel route and travel conditions for the user based on the user's health and emotional state.

[1053] "Emergency notification" refers to a system that detects abnormalities in a user's health condition in real time and promptly notifies care staff and medical institutions.

[1054] "Feedback" refers to users reporting the results of implementing a care plan and changes in their physical condition, providing data to improve the accuracy of the generative AI model based on this.

[1055] An "autonomous vehicle" is a vehicle that uses artificial intelligence to drive autonomously and reach its destination without the user having to operate it.

[1056] A specific system for implementing this invention is an application running in a self-driving vehicle that collects health and emotional data from users and provides optimal travel plans for elderly people based on that data.

[1057] Hardware and software used

[1058] Hardware:

[1059] Autonomous vehicle control unit

[1060] Smartwatches and fitness trackers worn by seniors

[1061] Cameras and microphones installed inside the vehicle

[1062] Sensor device (vehicle vital signs monitor)

[1063] software:

[1064] Generative AI models (e.g., deep learning models using TensorFlow or PyTorch)

[1065] Natural language processing technology (e.g., NLTK, spaCy)

[1066] Emotion engines (e.g. emotion recognition APIs such as Face++ and Affectiva)

[1067] What the program does

[1068] server:

[1069] The server periodically collects step counts, heart rate, and sleep data from smartwatches and fitness trackers.

[1070] It connects with medical institution databases to obtain users' medical records and health checkup results.

[1071] Emotion data is obtained from an emotion engine that uses a camera and microphone to recognize the user's face and analyze their tone of voice.

[1072] Data preprocessing:

[1073] The server removes noise from the collected data and corrects missing and outlier values.

[1074] Standardize the data and extract daily features (e.g., average daily steps, resting heart rate, total calorie intake, emotional data).

[1075] Generative AI models:

[1076] The generative AI model generates personalized care plans based on pre-processed data, while a model for predicting health needs is also built at the same time.

[1077] Emotional data is reflected in this model, enabling highly accurate predictions.

[1078] User Interface:

[1079] The generated care plan and predicted health needs are displayed on a display inside the vehicle.

[1080] Through this display, users can check their activity and health management instructions for the day.

[1081] Emergency Response:

[1082] The server monitors the user's current health data in real time.

[1083] If an abnormality is detected, the emergency notification system will be activated quickly to notify care staff and medical institutions.

[1084] Specific examples

[1085] 1. Data Collection and Analysis

[1086] Before an elderly person gets into an autonomous vehicle, their smartwatch sends their recent health data (step count, heart rate, and sleep data) to a server.

[1087] Cameras and microphones inside the vehicle collect user emotional data, which is then analyzed by an emotion engine.

[1088] 2. Generating a travel plan

[1089] Based on the collected data, the generative AI model suggests optimal travel routes for users (for example, routes that pass through parks or routes with less congestion).

[1090] The trip plan is displayed on the vehicle's display, allowing the user to view the suggested route and any important points to note.

[1091] 3. Emergency Response

[1092] Health data is monitored in real time during the journey, and if any abnormalities are detected, the vehicle will automatically stop and notify care staff or medical institutions.

[1093] Prompt Sentence Examples

[1094] Prompt: The user is currently a senior citizen. Suggest the best travel route based on his recent health and emotional data. He has low step counts and is feeling depressed, so suggest a relaxing route through the park.

[1095] This invention is a system that enables elderly people to travel safely and comfortably according to their individual health and emotional state. Real-time data monitoring and rapid emergency response provide an environment where autonomous vehicles can be used with greater peace of mind.

[1096] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1097] Step 1: Data collection

[1098] The server periodically collects step count, heart rate, and sleep data from the user's smartwatch or fitness tracker. It then connects with medical institution databases to obtain the user's medical records and health checkup results. It also uses the device's camera and microphone to analyze the user's facial recognition and tone of voice to assess their emotional state. This data is then input into the server, which outputs a comprehensive dataset on the user's health and emotional state based on the input data.

[1099] Step 2: Data Preprocessing

[1100] The server performs noise removal on the data collected in step 1 and corrects missing and outlier values. Next, it standardizes the data and extracts features such as average daily steps, resting heart rate, total calorie intake, and emotional data. It also converts text data (user feedback and medical records) into structured data using natural language processing technology. The standardized dataset is output.

[1101] Step 3: Generate an individualized care plan

[1102] The server generates a personalized care plan based on the pre-processed data using a generative AI model. The model is trained on a large dataset for elderly care. Based on the input data, the server outputs a care plan that includes daily activity instructions and recommended health behaviors.

[1103] Step 4: Reflecting emotional data

[1104] The server also incorporates emotional data into the generation of personalized care plans, providing highly accurate plans. Data obtained from the emotion engine is integrated with health data and analyzed using a generative AI model. A care plan corresponding to the emotional state based on the input data is output.

[1105] Step 5: View the care plan

[1106] The device then displays the generated care plan on the vehicle's internal display, where the user can view the day's activities and health management instructions. Based on the displayed data, the user can adjust their daily activities.

[1107] Step 6: Real-time health monitoring

[1108] As in step 1, the server monitors the user's health data in real time. If an abnormality is detected, an emergency notification system is activated immediately to notify care staff and medical institutions. This allows abnormality detection and response based on the input data.

[1109] Step 7: Optimize your travel plan

[1110] The server proposes an optimal travel plan based on the user's health and emotional state. It uses a generative AI model to calculate comfortable travel routes and travel conditions. Based on the input data, an optimized travel plan is output.

[1111] Step 8: Implement emergency response

[1112] The device detects abnormalities in the user's health condition in real time and sends an emergency notification immediately. The vehicle automatically stops and notifies pre-determined care staff and medical institutions. Emergency response is carried out based on the input data.

[1113] Step 9: Gather feedback

[1114] Users input the results of their care plan implementation and changes in their physical condition via their device. The server collects this feedback and uses it to retrain the generative AI model, improving the accuracy and effectiveness of the entire system. The retrained model is output based on the input data.

[1115] The specific processing unit 290 transmits the result of the specific processing to the smart glasses 214. In the smart glasses 214, the control unit 46A causes the speaker 240 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1116] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1117] In the above embodiment, an example in which the specific processing is performed by the data processing device 12 has been given, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the smart glasses 214.

[1118] [Third embodiment]

[1119] FIG. 5 shows an example of the configuration of a data processing system 310 according to the third embodiment.

[1120] 5, the data processing system 310 includes the data processing device 12 and a headset type terminal 314. An example of the data processing device 12 is a server.

[1121] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1122] The headset type terminal 314 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a display 343. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the display 343 are also connected to the bus 52.

[1123] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1124] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1125] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1126] Fig. 6 shows an example of the main functions of the data processing device 12 and the headset type terminal 314. As shown in Fig. 6, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1127] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1128] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1129] In the headset type terminal 314, a reception output process is performed by the processor 46. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1130] Next, a description will be given of the identification process performed by the identification processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as the "server" and the headset type terminal 314 will be referred to as the "terminal."

[1131] MODE FOR CARRYING OUT THE INVENTION

[1132] The present invention is a system for creating personalized care plans and predicting health needs using generative AI in Japan's elderly care service management system. Specific embodiments of the system are described below.

[1133] System Configuration

[1134] This system consists of users (elderly people or their caregivers), a server, and terminals (smartphones, tablets, PCs, etc.).

[1135] Program processing flow

[1136] 1. Data Collection

[1137] server

[1138] The server periodically collects step counts, heart rate, and sleep data from smartwatches and fitness trackers worn by users.

[1139] It connects with medical institution databases to obtain users' medical records and health checkup results.

[1140] It also collects daily activity logs, such as dietary habits, exercise, and medication intake, which are entered by the user via the device.

[1141] 2. Data Preprocessing

[1142] server

[1143] Noise is removed from the collected data and missing values ​​are corrected.

[1144] The total number of steps taken each day is calculated based on the step count data from the smartwatch, and the resting heart rate is calculated based on the heart rate data.

[1145] Text data (e.g., meal details) is converted into structured data using natural language processing technology.

[1146] 3. Model Building

[1147] server

[1148] A generative AI model is trained using large amounts of pre-collected elderly care and medical data, and the model uses deep learning techniques.

[1149] The model generates personalized care plans and predicts future health needs based on the user's features.

[1150] 4. Generate personalized care plans

[1151] User

[1152] Users enter basic information (age, gender, medical history) and daily activity data through the device.

[1153] The server uses this information to generate an individualized care plan using a generative AI model.

[1154] The generated care plan is displayed on the device, allowing the user to check daily activity plans and recommended health behaviors.

[1155] 5. Anticipating health needs

[1156] server

[1157] The server uses a generative AI model to predict future health risks (e.g., risk of developing nutritional deficiencies or high blood pressure) based on the user's current health and lifestyle data.

[1158] The prediction results are communicated to the user and care staff in real time.

[1159] 6. Collaboration with care staff

[1160] Terminal

[1161] The individualized care plan and health needs prediction results received from the server are sent to the care staff's terminal.

[1162] The care staff adjusts specific care responses for the user based on the information received.

[1163] 7. Feedback Loops

[1164] User

[1165] The user provides feedback to the server via their device about the results of the care plan, their impressions, and changes in their physical condition.

[1166] The server analyzes the collected feedback and uses it to retrain the generative AI model.

[1167] Specific examples

[1168] Example 1: Data collection and preprocessing

[1169] The server automatically collects nighttime sleep data and daytime step count data from the smartwatch worn by the user every day, and cleanses this data to extract useful features.

[1170] Example 2: Generating personalized care plans

[1171] When a user enters their regular health check data into the terminal, the server uses that data to generate an individualized care plan, such as "walking 8,000 steps per day" or "light aerobic exercise twice a week," and displays it to the user.

[1172] Example 3: Predicting health needs

[1173] The server analyzes data entered by the user, such as a recent loss of appetite, and evaluates the estimated risk of nutritional deficiency. The generative AI model then suggests taking a multivitamin supplement and sends an alert to care staff.

[1174] The system allows users to implement personalized care plans to maintain and improve their health, and it also allows care staff to respond more accurately, improving the quality of elderly care.

[1175] The processing flow will be explained below.

[1176] Step 1: Data collection

[1177] server

[1178] The server works in conjunction with smartwatches and fitness trackers to periodically collect the user's steps, heart rate, and sleep data.

[1179] It connects to medical institution databases via API to obtain users' past medical records and regular health checkup results.

[1180] Users input their daily activities and dietary information into the device, and this data is also sent to the server.

[1181] Step 2: Data Preprocessing

[1182] server

[1183] The server cleanses the collected data, corrects missing and outlier values, and removes inappropriate data.

[1184] Standardize the data and extract features (e.g., average daily steps, resting heart rate, total calorie intake).

[1185] Text data (meal details, user feedback) is converted into structured data using natural language processing technology.

[1186] Step 3: Model Building

[1187] server

[1188] The server trains generative AI models using large datasets for elderly care.

[1189] Deep learning algorithms are used for training to build models that generate personalized care plans from user features and predict health needs.

[1190] Evaluate the accuracy of the model and tune the parameters as needed.

[1191] Step 4: Generate an individualized care plan

[1192] User

[1193] Users enter their basic information (age, gender, medical history) and daily activity data through the device.

[1194] server

[1195] The server collects the information submitted by the user and generates a personalized care plan using a generative AI model.

[1196] The generated care plan includes daily activity instructions and recommended health behaviors (e.g., aiming for 5,000 steps per day, eating a balanced diet).

[1197] Step 5: View Care Plan

[1198] Terminal

[1199] The designated care plan is displayed on the device screen, allowing the user to check the day's activities and health management instructions.

[1200] Step 6: Anticipate health needs

[1201] server

[1202] By analyzing a user's current health and lifestyle data, a generative AI model predicts future health risks (e.g., nutritional deficiencies, risk of chronic diseases).

[1203] Based on the prediction results, the system suggests necessary health management actions to the user (e.g., taking specific supplements, undergoing regular medical checkups).

[1204] Step 7: Working with notifications

[1205] Terminal

[1206] The prediction results and individualized care plan received from the server are sent to the care staff's terminal.

[1207] Care staff

[1208] Based on the information received, the care staff plans and implements specific care responses for the user.

[1209] Step 8: Gather feedback

[1210] User

[1211] Users input the results of the care plan, their impressions, and changes in their physical condition through their terminals, and provide feedback to the server.

[1212] server

[1213] The collected feedback is analyzed and used to improve and retrain the generative AI model.

[1214] Through feedback loops, the accuracy and effectiveness of the model and the system as a whole are continually improved.

[1215] Example 1

[1216] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1217] To improve the quality of elderly care, it is important to accurately generate individualized care plans and predict future health risks. However, conventional systems have difficulty centrally managing and effectively analyzing a user's diverse health data. Furthermore, they lack sufficient collaboration with care staff, making it difficult to provide appropriate care. Furthermore, they lack a mechanism for appropriately incorporating user feedback and updating the model, making it difficult to respond to ever-changing health conditions.

[1218] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1219] In this invention, the server includes: means for collecting health data from a user's activity tracking device; means for acquiring medical data in cooperation with a medical institution's database; means for collecting daily activity data entered by the user; means for performing noise reduction and missing value correction on the collected health data; means for calculating the total number of steps per day based on step count data and the resting heart rate based on heart rate data; means for converting text data into structured data using natural language processing technology; means for training a generative AI model using elderly care data and medical data; means for generating an individualized care plan using the generative AI model; means for displaying the generated care plan; means for predicting future health needs using the generative AI model; means for notifying care staff of the prediction results and the care plan; means for collecting user feedback and using it to retrain the model; means for care staff to adjust care responses to the user based on the care plan and prediction results received; and means for the user to provide feedback. This enables the generation of individualized care plans and prediction of health risks, improving the quality of elderly care. Furthermore, by appropriately incorporating user feedback and updating the model, care can be provided that is always tailored to the latest health status.

[1220] "User" refers to the elderly person or their caregiver who uses the system.

[1221] An "activity measuring device" is a device for collecting a user's physical activity data, and includes smart watches, fitness trackers, etc.

[1222] "Health data" refers to information such as a user's number of steps, heart rate, sleep data, medical records, health checkup results, dietary habits, exercise amount, and medication intake.

[1223] "Server" refers to the computer system that collects, preprocesses, analyzes, and stores user data, and trains and runs generative AI models.

[1224] "Noise reduction" refers to the process of removing unnecessary or erroneous information from collected data.

[1225] "Missing value correction" refers to the process of making predictions or inferences to fill in missing values ​​in data.

[1226] "Natural language processing technology" refers to technology for converting text data into a format that is easy for machines to understand.

[1227] "Generative AI model" refers to an AI (artificial intelligence) model that uses large amounts of data to generate personalized care plans and predict future health needs.

[1228] "Individualized care plan" refers to a care plan created based on a user's specific health condition and lifestyle habits.

[1229] "Health needs prediction" refers to the process of predicting future health risks and care needs based on a user's current data.

[1230] "Care staff" refers to professional people who provide care for older people.

[1231] "Feedback" refers to information provided by the user regarding the results of implementing the care plan, their impressions, and changes in their physical condition.

[1232] "Model retraining" refers to the process of updating a generative AI model with newly collected data and feedback to improve its accuracy.

[1233] This invention is a system for creating personalized care plans and predicting health needs using a generative AI model in an elderly care service management system in Japan. The system consists of users (elderly people or their caregivers), a server, and devices (smartphones, tablets, personal computers, etc.).

[1234] System Configuration

[1235] The system is configured as follows:

[1236] server

[1237] The server is responsible for:

[1238] 1. Periodically collect step count, heart rate, and sleep data from the user's activity tracking device (smartwatch or fitness tracker).

[1239] 2. Link with medical institution databases to obtain users' medical records and health checkup results.

[1240] 3. It also collects daily activity logs, such as dietary details, exercise volume, and medication intake, entered by the user through the device.

[1241] 4. Remove noise from the collected data and correct missing values.

[1242] 5. Calculate the total number of steps taken each day based on the step count data from the smartwatch, and calculate the resting heart rate based on the heart rate data.

[1243] 6. Convert text data (e.g., meal details) into structured data using natural language processing technology.

[1244] 7. Train generative AI models using elderly care and medical data, which use deep learning techniques.

[1245] 8. The model generates a personalized care plan and predicts future health needs based on the user's features.

[1246] Terminal

[1247] The device is used by users and care staff and has the following functions:

[1248] 1. The user enters basic information (age, gender, medical history) and daily activity data through the device.

[1249] 2. The device displays the personalized care plan generated using the generative AI model.

[1250] 3. Notify users and care staff in real time of health risk prediction results and the implementation results of care plans.

[1251] 4. Care staff will coordinate specific care responses for the user based on the information received.

[1252] User

[1253] Users provide daily activity data and feedback to the system:

[1254] 1. Users send health data to a server via their smartwatch or fitness tracker.

[1255] 2. The user enters information such as dietary habits and medication intake into the terminal application.

[1256] 3. Provide feedback on the results of the care plan, impressions, and changes in physical condition via the device application.

[1257] Specific examples

[1258] Example 1: Data collection and preprocessing

[1259] The server automatically collects nighttime sleep data and daytime step count data from the smartwatch worn by the user every day, and cleanses this data to extract useful features.

[1260] Example 2: Generating personalized care plans

[1261] When a user enters their regular health check data into the terminal, the server uses that data to generate an individualized care plan, such as "walking 8,000 steps per day" or "light aerobic exercise twice a week," and displays it to the user.

[1262] Example 3: Predicting health needs

[1263] The server analyzes data entered by the user, such as a recent loss of appetite, and assesses the estimated risk of nutritional deficiency. The generative AI model then suggests taking a multivitamin supplement and sends an alert to care staff. This system allows users to implement personalized care plans to maintain and improve their health. It also allows care staff to respond more accurately, improving the quality of elderly care.

[1264] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1265] Step 1: Data collection

[1266] Server collects smartwatch data

[1267] Input: Step count, heart rate, and sleep data from the user's smartwatch

[1268] Data processing: Collect data from the smartwatch periodically and store it in a database

[1269] Output: Raw health data

[1270] Specific operation: When the user syncs the smartwatch every morning, the server receives the night's sleep data and the previous day's step count data and stores them in a database.

[1271] Acquisition of medical data by the server

[1272] Input: Medical records and health check results from medical institution databases

[1273] Data processing: Data is acquired through the medical institution's API and saved as user health information.

[1274] Output: Retrieved medical data

[1275] Specific operation: Outside of the medical institution's scheduled maintenance period, the server sends an API request to obtain the user's latest blood test results and diagnostic data.

[1276] Manual data collection by the server

[1277] Input: Dietary information, exercise amount, and medication intake status entered by the user via the device

[1278] Data processing: Collect input data in real time and save it in a database

[1279] Output: Daily activity data

[1280] Specific operation: After dinner, the user uses the terminal application to input the details of their meal and medication intake, and the data is sent to the server in real time.

[1281] Step 2: Data Preprocessing

[1282] Server-based noise removal and missing value correction

[1283] Input: Collected health data

[1284] Data processing: data denoising and missing value imputation

[1285] Output: Cleaned health data

[1286] Specific operation: The server cleanses the data obtained from the smartwatch and fills in missing daily step count data with data from the previous day.

[1287] Server recalculation and structuring of data

[1288] Input: Data after noise removal and correction

[1289] Data processing: Counting step count data, calculating resting heart rate, structuring text data

[1290] Output: Recalculated and structured data

[1291] Specific operation: The server aggregates the step count data and calculates the total number of steps taken for the day. It also analyzes the text of meal descriptions entered by the user, such as "salad and grilled chicken," and extracts information about calories and nutrients as structured data.

[1292] Step 3: Model Building

[1293] Server-based training of generative AI models

[1294] Input: Elderly care data, medical data

[1295] Data processing: Using deep learning techniques to train AI models

[1296] Output: A trained generative AI model

[1297] How it works: Generative AI models are trained using historical medical datasets and iteratively trained until error is minimized.

[1298] Step 4: Generate an individualized care plan

[1299] User enters basic information

[1300] Input: User's basic information (age, gender, medical history), daily activity data

[1301] Data processing: Enter basic information into the system and send it to the server

[1302] Output: User basic information data

[1303] Specific operation: A user logs in to the application and enters their age, gender, and medical history on the profile screen.

[1304] Server-based care plan generation

[1305] Input: User's basic information and daily activity data

[1306] Data processing: Generative AI models are used to generate personalized care plans

[1307] Output: Individualized Care Plan

[1308] Specific operation: The server calls up an AI model based on the basic information entered and generates a daily activity schedule, recommended exercises, and a meal plan.

[1309] Displaying care plans on devices

[1310] Input: Generated personalized care plan

[1311] Data processing: Push notification of care plan to device

[1312] Output: The care plan displayed to the user

[1313] What it does: The care plan is pushed to the device, and when the user opens the app, a detailed activity plan is displayed on the dashboard.

[1314] Step 5: Anticipate health needs

[1315] Server-based health risk prediction

[1316] Input: User's current health and lifestyle data

[1317] Data processing: Predicting future health risks using generative AI models

[1318] Output: Health risk prediction results

[1319] How it works: Analyzes recent dietary data to assess the user's risk of vitamin deficiency. If the deficiency persists, the AI ​​model predicts the risk and generates an alert.

[1320] Real-time notifications

[1321] Input: Health risk prediction results

[1322] Data processing: Real-time notification of prediction results

[1323] Output: Notification to user and care staff

[1324] Specific operation: A push notification is sent to the care staff's device, and an alert is displayed stating, "User A is at risk of high blood pressure."

[1325] Step 6: Collaborate with care staff

[1326] Sending care information by device

[1327] Input: Personalized care plan and health risk prediction results received from the server

[1328] Data processing: Sending information to care staff's terminal

[1329] Output: Care plan and health risk prediction information

[1330] Specific operation: Information from the server is sent to an app dedicated to care staff, who can check it in real time.

[1331] Care staff coordination

[1332] Input: Care plan and health risk prediction information

[1333] Data processing: Coordination of care responses

[1334] Output: Adjusted care plan

[1335] Specific operation: During regular visits, care staff use information obtained from the server to fine-tune the user's care plan.

[1336] Step 7: Feedback Loop

[1337] User feedback

[1338] Input: Care plan implementation results, impressions, changes in physical condition

[1339] Data processing: Input the feedback into the system and send it to the server

[1340] Output: User feedback information

[1341] Specific operation: When the user enters feedback into the app, such as "I've been walking for a while, and as a result, I feel better," that information is sent to the server.

[1342] Server data reuse

[1343] Input: Collected feedback data

[1344] Data processing: Retraining generative AI models

[1345] Output: Improved generative AI model

[1346] Specific operation: The server analyzes the collected feedback data and performs re-training to improve the accuracy of the AI ​​model.

[1347] (Application example 1)

[1348] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1349] In conventional elderly care services, collecting health data and creating care plans is often done manually, making it difficult to provide personalized care. Furthermore, there is a lack of means to monitor health conditions in real time and quickly respond as needed. This poses a particular challenge when an elderly person becomes ill while traveling or when emergency care is required.

[1350] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1351] In this invention, the server includes means for collecting user health data, means for preprocessing the collected health data, means for generating an individualized care plan based on the preprocessed data using a generative AI model, means for displaying the generated care plan in the vehicle, means for predicting future health needs using the generative AI model, means for notifying a caregiver of the prediction results and the care plan, means for collecting feedback from the user and using it to retrain the model, means for optimizing a driving route according to the user's health condition, and means for detecting and notifying abnormal health data in real time, thereby enabling the provision of an individualized care plan, real-time health condition monitoring, and appropriate responses even while traveling.

[1352] "User" refers to an individual about whom health data is collected, particularly the elderly and those in need of health management.

[1353] "Health data" refers to information that indicates an individual's health status, such as heart rate, number of steps, sleep data, dietary content, amount of exercise, and medication intake.

[1354] "Preprocessing" refers to the processing of collected health data, such as noise removal, missing value correction, and data structuring.

[1355] A "generative AI model" is an artificial intelligence model trained using deep learning techniques to generate personalized care plans based on collected health data and predict future health needs.

[1356] A "personalized care plan" is a specific action plan created based on the user's individual health data and lifestyle, with the aim of maintaining or improving the health of a specific user.

[1357] An "autonomous vehicle" is a vehicle that is capable of driving independently and is equipped with the functionality to collect and process health data.

[1358] "Display means" refers to a device or function for visually showing the generated care plan to the user or caregiver (e.g., an in-car display or a smartphone).

[1359] "Health needs" are the activities and care requirements necessary to maintain or improve the user's future health status.

[1360] "Care aides" are professionals and caregivers who provide care to older people and people with health care needs.

[1361] "Notification means" refers to a device or system that notifies caregivers of the generated care plan and the predicted health needs (e.g., a smartphone app or an in-car system).

[1362] "Feedback" is information that the user reports, such as the results of implementing the care plan, their impressions, and changes in their physical condition.

[1363] "Model retraining" is a machine learning process that uses collected feedback data to further improve a generative AI model.

[1364] "Route optimization" is the process of selecting the optimal driving route and rest points based on the user's health condition.

[1365] "Real-time detection" is a function that can immediately recognize and respond to abnormalities in health data when they occur.

[1366] The present invention relates to a system for providing personalized health care in autonomous vehicles, particularly designed for the elderly and people with health care needs. Specific embodiments of the system are described below.

[1367] System Configuration

[1368] This system consists of a user (such as an elderly person), a server (the autonomous vehicle's infotainment system), a device (such as a smartwatch or fitness tracker) for collecting health data, and a caregiver (such as a caregiver or medical staff).

[1369] Program processing flow

[1370] Data collection

[1371] The server collects real-time health data such as heart rate, step count, and sleep data from the smartwatch or fitness tracker worn by the user. It also connects to medical institution databases to obtain the user's medical records and health checkup results. It also collects dietary information and daily activity logs entered by the user through a terminal inside the autonomous vehicle.

[1372] Data Preprocessing

[1373] The server removes noise from the collected health data and corrects missing values. It calculates the total number of steps taken each day based on the step count data from the smartwatch and calculates the resting heart rate based on the heart rate data. Text data such as meal details is converted into structured data using natural language processing technology.

[1374] Model Building

[1375] The server uses large amounts of pre-collected elderly care and medical data to train a generative AI model, which uses deep learning techniques to generate personalized care plans and predict future health needs based on the user's features.

[1376] Personalized care plan generation

[1377] Users enter basic information (age, gender, medical history) and daily activity data through the infotainment system. The server uses this information to generate a personalized care plan using a generative AI model. The generated care plan is displayed on a display inside the autonomous vehicle, allowing users to check their daily activity plan and recommended health behaviors.

[1378] Anticipating health needs

[1379] The server uses a generative AI model to predict future health risks (e.g., risk of developing nutritional deficiencies or high blood pressure) based on the user's current health and lifestyle data. The prediction results are notified to the user and caregivers in real time.

[1380] Cooperation with care staff

[1381] The personalized care plan and health needs prediction results received from the server are sent to the caregiver's device, who then adjusts specific care responses for the user based on the received information.

[1382] Feedback Loop

[1383] Users provide feedback to the server via their devices about the results of implementing the care plan, their impressions, and changes in their physical condition. The server analyzes the collected feedback and uses it to retrain the generative AI model.

[1384] Hardware and software used

[1385] Smartwatch: Collects heart rate, steps, and sleep data.

[1386] Autonomous vehicles: Infotainment systems for collecting and processing health data.

[1387] Server: Preprocessing data, training and running generative AI models.

[1388] Deep learning framework (TensorFlow or PyTorch): Implementing and training generative AI models.

[1389] Natural language processing technologies (SpaCy, NLTK, etc.): Structuring text data.

[1390] Specific examples

[1391] For example, if a user's heart rate increases while riding in an autonomous vehicle, the autonomous vehicle's system will collect heart rate data in real time and provide a care plan to "take a deep breath and relax" and a notification recommending a break at the next service area.

[1392] Prompt Sentence Examples

[1393] "The system monitors passengers' health status in real time based on heart rate and step count data during the ride, and immediately proposes an appropriate care plan if an abnormality is detected. It also suggests optimal routes and rest points depending on the passenger's condition."

[1394] The system enables personalized care plans, real-time health monitoring and appropriate responses, even while on the move.

[1395] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1396] Step 1:

[1397] A user wears a smartwatch or fitness tracker and gets into an autonomous vehicle. The device collects real-time heart rate, step count, and sleep data from the smartwatch and sends them to a server. The input of this step is the health data from the smartwatch, and the output is the raw data sent to the server.

[1398] Step 2:

[1399] The server preprocesses the received health data. Specifically, it removes noise and corrects missing values. For example, it complements missing data based on past data and removes obviously erroneous data. The input is the raw data collected in step 1, and the output is the cleansed health data.

[1400] Step 3:

[1401] The server uses a generative AI model to generate a personalized care plan based on the preprocessed data. For example, it may recommend deep breathing if the heart rate is high or light exercise if the step count is low. The input for this step is the preprocessed health data, and the output is a personalized care plan.

[1402] Step 4:

[1403] The generated care plan is presented to the user through the autonomous vehicle's infotainment system, where the user can visually review it and follow the instructions. The input is the generated personalized care plan, and the output is the care plan displayed to the user.

[1404] Step 5:

[1405] The server uses the generative AI model to predict the user's future health needs. For example, it evaluates the risk of rising blood pressure several months from now based on current data. The input of this step is the preprocessed health data, and the output is the predicted health risks and future health needs.

[1406] Step 6:

[1407] The server notifies the caregiver of the prediction results and the care plan. The caregiver can then adjust specific care responses for the user based on this information. The input is the prediction results and the care plan, and the output is a notification sent to the caregiver.

[1408] Step 7:

[1409] The server provides an interface for collecting feedback from users. Users input the results of implementing the care plan, their impressions, and changes in their physical condition. The input is the user's feedback, and the output is the feedback data stored on the server.

[1410] Step 8:

[1411] The server analyzes the collected feedback and uses it to retrain the generative AI model, which enables the model to provide more accurate care plans. The input is the feedback data, and the output is the retrained generative AI model.

[1412] Step 9:

[1413] The server optimizes the driving route according to the user's health condition. For example, it suggests appropriate rest areas according to the user's fatigue level. The input is health data detected in real time, and the output is the optimized driving route.

[1414] Step 10:

[1415] The server detects abnormal health data in real time and immediately notifies the user and caregivers, for example, sending an alert if the heart rate is abnormally high. The input is the health data collected in real time, and the output is the alert notification sent.

[1416] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1417] MODE FOR CARRYING OUT THE INVENTION

[1418] The present invention is a system for creating personalized care plans and predicting health needs using generative AI and an emotion engine in an elderly care service management system in Japan. Specific embodiments of the system are described below.

[1419] System Configuration

[1420] This system consists of users (elderly people or their caregivers), a server, terminals (smartphones, tablets, PCs, etc.), and an emotion engine.

[1421] Program processing flow

[1422] 1. Data Collection

[1423] server

[1424] The server periodically collects step counts, heart rate, and sleep data from smartwatches and fitness trackers worn by users.

[1425] It connects with medical institution databases to obtain users' medical records and health checkup results.

[1426] It also collects daily activity logs, such as dietary habits, exercise, and medication intake, which are entered by the user via the device.

[1427] 2. Collecting Emotional Data

[1428] Emotion Engine

[1429] The emotion engine uses facial recognition technology to analyze the user's facial expressions and assess their emotional state (e.g., happy, sad, or angry).

[1430] Analyze the tone of the user's voice to recognize their emotional state.

[1431] 3. Data Preprocessing

[1432] server

[1433] Noise is removed from the collected health and emotion data, missing and outliers are corrected, and inappropriate data is removed.

[1434] The data is standardized and features (e.g., average daily steps, resting heart rate, total calorie intake, and emotional data) are extracted.

[1435] Text data (meal details, user feedback) is converted into structured data using natural language processing technology.

[1436] 4. Model Building

[1437] server

[1438] A generative AI model is trained using a deep learning algorithm on a large dataset for elderly care.

[1439] The model builds a model that generates an individualized care plan based on the user's features and a model that predicts health needs.

[1440] It also takes into account sentiment data, evaluates the accuracy of the model, and tunes parameters as needed.

[1441] 5. Generate personalized care plans

[1442] User

[1443] Users enter basic information (age, gender, medical history) and daily activity data through the device.

[1444] server

[1445] The server collects the emotion data from the emotion engine and the information submitted by the user, and generates a personalized care plan using a generative AI model.

[1446] The generated care plan includes daily activity instructions and recommended health behaviors (e.g., aiming for 5,000 steps per day, eating a balanced diet).

[1447] 6. Care plan display

[1448] Terminal

[1449] The care plan is displayed on the device screen, allowing users to check their activities and health management instructions for the day.

[1450] 7. Anticipating health needs

[1451] server

[1452] By analyzing a user's current health, lifestyle, and emotional data, a generative AI model predicts future health risks (e.g., risk of nutritional deficiencies and high blood pressure).

[1453] Based on the prediction results, the system suggests necessary health management actions to the user (e.g., taking specific supplements, undergoing regular medical checkups).

[1454] 8. Notifications and Collaboration

[1455] Terminal

[1456] The prediction results and individualized care plan received from the server are sent to the care staff's terminal.

[1457] Care staff

[1458] Based on the information received, the care staff plans and implements specific care responses for the user.

[1459] 9. Feedback Collection

[1460] User

[1461] Users input the results of the care plan, their impressions, and changes in their physical condition through their terminals, and provide feedback to the server.

[1462] server

[1463] The collected feedback is analyzed and used to improve and retrain the generative AI model.

[1464] Through feedback loops, the accuracy and effectiveness of the model and the system as a whole are continually improved.

[1465] Specific examples

[1466] Example 1: Data collection and preprocessing

[1467] The user collects daily step counts and heart rate data from the smartwatch, and the server cleans the data to extract useful features. At the same time, the emotion engine collects the user's emotional data using the device's camera and microphone.

[1468] Example 2: Generating personalized care plans

[1469] Users who tend to be isolated input data from regular health checkups and daily life into the device, and the server generates an individualized care plan based on that data and emotional data obtained from the emotion engine, including recommendations such as "walking 7,000 steps a day" and "communicating with friends three times a week," and displays it to the user.

[1470] Example 3: Predicting health needs

[1471] The server analyzes data showing that the user enjoys walking but has recently been feeling depressed, and the generative AI model predicts that the user has a tendency toward seasonal depression. Reflecting the data from the emotion engine, the server recommends the user engage in daily light exercise and sun exposure, and notifies care staff to contact the user regularly.

[1472] The system allows users to receive scientifically-backed, personalized care plans and responses tailored to their emotional state, helping them maintain and improve their physical and mental health, while providing caregivers with more accurate information to provide specific care.

[1473] The processing flow will be explained below.

[1474] Step 1: Data collection

[1475] server

[1476] The server works in conjunction with smartwatches and fitness trackers to periodically collect the user's steps, heart rate, and sleep data.

[1477] It connects to medical institution databases via API to obtain users' past medical records and regular health checkup results.

[1478] Daily activity logs, such as what the user eats, how much exercise they do, and what medications they take, are also sent to the server.

[1479] Step 2: Collecting emotion data

[1480] Emotion Engine

[1481] The emotion engine uses the device's camera to capture the user's facial expressions and analyzes their emotional state (happiness, sadness, anger, etc.) using facial recognition technology.

[1482] When a user speaks using the device, the tone of their voice is analyzed through a microphone to recognize their emotional state.

[1483] Step 3: Data Preprocessing

[1484] server

[1485] The server cleanses the collected health and emotion data and corrects missing or outlier values.

[1486] The data is standardized and features (e.g., average daily steps, resting heart rate, total calorie intake, and emotional data) are extracted.

[1487] Text data (meal details, user feedback) is converted into structured data using natural language processing technology.

[1488] Step 4: Model Building

[1489] server

[1490] A generative AI model is trained using a large dataset for elderly care, using deep learning algorithms.

[1491] The model will be constructed to generate an individualized care plan based on the user's features, and to predict health needs.

[1492] The accuracy of the model is evaluated taking into account sentiment data and parameters are adjusted as needed.

[1493] Step 5: Generate an individualized care plan

[1494] User

[1495] Users enter their basic information (age, gender, medical history) and daily activity data into the device.

[1496] server

[1497] The server collects the emotion data from the emotion engine and the information submitted by the user, and generates a personalized care plan using a generative AI model.

[1498] The user's plan includes daily activity instructions and recommended health behaviors (e.g., aiming for 5,000 steps per day, adjusting diet, and taking mental health measures).

[1499] Step 6: View Care Plan

[1500] Terminal

[1501] The generated care plan is displayed on the terminal screen, and the user can check the day's activities and health management instructions.

[1502] Step 7: Anticipate health needs

[1503] server

[1504] The server analyzes the user's current health, lifestyle, and emotional data, and uses a generative AI model to predict future health risks (e.g., risk of nutritional deficiencies and high blood pressure).

[1505] Based on the prediction results, the system suggests necessary health management actions to the user (e.g., taking specific supplements, undergoing regular medical checkups).

[1506] Step 8: Working with notifications

[1507] Terminal

[1508] The prediction results and individualized care plan received from the server are sent to the care staff's terminal.

[1509] Care staff

[1510] Based on the information received, the care staff plans and implements specific care responses for the user.

[1511] Step 9: Gather feedback

[1512] User

[1513] The user inputs the results of the care plan, their impressions, and any changes in their physical condition into the terminal and provides feedback to the server.

[1514] server

[1515] The collected feedback is analyzed and used to improve and retrain the generative AI model.

[1516] Through feedback loops, the accuracy and effectiveness of the model and the system as a whole are continually improved.

[1517] Example 2

[1518] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1519] In elderly care services, it is challenging to provide care plans that take into account the user's individual health and emotional state. While conventional systems focus on collecting and analyzing health data, incorporating the user's emotional data is needed to provide more accurate personalized care plans. However, such systems face challenges due to the complexity of data collection and the need to assess the user's emotional state in real time. Furthermore, a means of clearly communicating the generated care plan and health risk predictions to the user is also required, and a mechanism for seamlessly aggregating subsequent feedback and retraining the model is lacking.

[1520] The identification processing by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting user health data, means for preprocessing the collected health data, means for generating an individualized care plan using a generative AI model based on the preprocessed data, means for collecting user emotion data using facial recognition technology and voice analysis technology, means for preprocessing the emotion data, means for generating an individualized care plan using the preprocessed emotion data and health data, means for displaying the generated care plan, means for predicting future health needs using the generative AI model, means for notifying care staff of the prediction results and the care plan, and means for collecting feedback from the user and using it to retrain the model. This makes it possible to provide an accurate individualized care plan based on the user's health condition and emotional state, and is expected to improve the accuracy of health risk predictions and the quality of care.

[1521] A "user" is an entity that utilizes the system to provide health and emotional data and receive a personalized care plan.

[1522] "Health data" refers to data collected by users in their daily lives, such as the number of steps taken, heart rate, sleep data, dietary content, amount of exercise, and medication intake.

[1523] "Emotional data" refers to data relating to the emotional state of a user obtained from facial expressions and tone of voice collected using face recognition technology and voice analysis technology.

[1524] "Facial recognition technology" is a technology that uses a camera to analyze a user's facial expressions and assess their emotional state.

[1525] "Voice analysis technology" is a technology that uses a microphone to analyze the tone and intonation of a user's voice to assess their emotional state.

[1526] A "generative AI model" is an algorithmic model that uses collected health and emotional data to generate personalized care plans and predict future health needs.

[1527] A "personalized care plan" is a plan that includes health care instructions and activity recommendations that are optimal for a user, generated based on the user's health and emotional data.

[1528] "Preprocessing" refers to processing of collected data, such as noise removal, missing value correction, outlier removal, data standardization, and feature extraction.

[1529] "Feedback" is information provided by the user regarding the results of implementing the care plan, their impressions, and changes in their physical condition, and is used to retrain the model.

[1530] "Care staff" are people who plan and implement specific care responses for users based on the individualized care plans and health needs prediction results generated using the system.

[1531] "Notification" refers to the act of notifying the user or care staff of the generated care plan or prediction results via a terminal.

[1532] MODE FOR CARRYING OUT THE INVENTION

[1533] This invention is a system for creating personalized care plans and predicting health needs in elderly care services using generative AI and an emotion engine. This system consists of a user, a server, a terminal, and an emotion engine.

[1534] 1. Users

[1535] Users wear a smartwatch or fitness tracker to record their daily activity data. They also input their daily activity log, including their diet, exercise, and medication intake, via a device (smartphone, tablet, PC, etc.). Furthermore, the device's camera and microphone are used to record data on their emotional state.

[1536] 2. Server

[1537] The server does the following:

[1538] Data collection: The server periodically collects data from the user's smartwatch or fitness tracker and connects with medical institution databases to obtain medical records and health checkup results.

[1539] Emotional data collection: We collect user emotional data using facial recognition and voice analysis technologies.

[1540] Data preprocessing: Noise removal, missing value correction, outlier removal, standardization, feature extraction, and text data structuring are performed on the collected health and emotion data.

[1541] Model Building: Train generative AI models using large-scale elderly care datasets to build and evaluate models that generate personalized care plans and predict health needs.

[1542] Prediction and generation: Using a generative AI model, future health risks are predicted based on the user's features and emotional data, and an individualized care plan is generated.

[1543] Notification and collaboration: Prediction results and care plans are notified to care staff devices.

[1544] 3. Terminal

[1545] The terminal has the following features:

[1546] Feedback collection: Users input the results of the care plan, their impressions, and changes in their physical condition via their terminal and provide feedback to the server.

[1547] Care plan display: The generated care plan is displayed on the terminal screen, allowing the user to check the health care instructions for the day.

[1548] Notification: Display the prediction results and care plan sent from the server and notify the care staff.

[1549] Specific examples

[1550] Example 1: Data collection and preprocessing

[1551] The user records their daily steps and heart rate on their smartwatch, and the server cleans and extracts features from the data. The emotion engine collects the user's emotion data using the device's camera and microphone.

[1552] Example 2: Generating personalized care plans

[1553] Users who tend to be isolated input data from regular health checkups and daily life into the device, and the server generates an individualized care plan based on that data and emotional data obtained from the emotion engine, including recommendations such as "walking 7,000 steps a day" and "communicating with friends three times a week," and displays it to the user.

[1554] Example 3: Predicting health needs

[1555] The server analyzes data showing that the user enjoys walking but has recently been feeling depressed, and the generative AI model predicts that the user has a tendency toward seasonal depression. Reflecting the data from the emotion engine, the server recommends the user engage in daily light exercise and sun exposure, and notifies care staff to contact the user regularly.

[1556] Prompt Sentence Examples

[1557] "The user is 70 years old, their primary daily activity is walking, and the emotion engine has detected their recent emotional state as 'sad'. Please generate a personalized care plan for this user."

[1558] The flow of the identification process in the second embodiment will be described with reference to FIG.

[1559] System program processing flow

[1560] Step 1: Data collection

[1561] The server periodically collects data such as step counts, heart rate, and sleep data from the smartwatch or fitness tracker worn by the user. As input, it receives biometric data from the smartwatch or fitness tracker. Specifically, it makes API calls from the smart device and stores the data in the server's database.

[1562] The server connects to the medical institution's database to retrieve the user's medical records and health checkup results. It receives data obtained from the medical institution's API as input. Specifically, it calls the API on a regular schedule, retrieves the necessary data, and stores it in the server's database.

[1563] Users input their daily activity logs, including their diet, exercise, and medication intake, via a device. The system receives inputs from users' manual entries and log data written on a daily basis. Specifically, data is entered through a dedicated application on the device and sent to a server. The output is the collected health data and daily activity log.

[1564] Step 2: Collecting emotion data

[1565] The emotion engine uses facial recognition technology to analyze video data acquired from the device's camera and analyze the user's facial expressions. It receives the video data acquired from the camera as input. Specifically, it uses a facial recognition algorithm to analyze the facial expressions and evaluate the user's emotional state, such as "happiness" or "sadness." The output is data indicating the emotional state.

[1566] The emotion engine analyzes audio data recorded using a microphone and recognizes emotions based on the user's tone of voice. It receives audio data from the microphone as input, analyzes the audio data using a voice analysis algorithm, and evaluates the user's emotional state. The output is audio data that indicates the user's emotional state.

[1567] Step 3: Data Preprocessing

[1568] The server performs noise removal on the collected health and emotion data. As input, it receives the collected raw data. Specifically, it detects outliers and missing values ​​and removes or corrects them. The output is the cleansed data.

[1569] The server normalizes the data and extracts features such as average daily steps, resting heart rate, total calorie intake, and emotional data. It receives the cleansed data as input and applies data normalization and feature extraction algorithms to convert it into analyzable data. The output is feature data for analysis.

[1570] The server converts text data (meal details, user feedback) into structured data using natural language processing technology. It receives text data entered by the user as input. Specifically, it applies a text analysis algorithm to convert the data into structured data, such as meal items. The output is structured text data.

[1571] Step 4: Model Building

[1572] The server trains a generative AI model using a large-scale dataset for elderly care. It receives a large amount of training data as input. Specifically, it uses a deep learning algorithm (e.g., TensorFlow or PyTorch) to train the model. The output is a trained generative AI model.

[1573] The server builds a model that generates an individualized care plan and a model that predicts health needs based on the user's features. It receives feature data as input. Specifically, it uses the generative AI model to create an optimal care plan and predict health risks for each user. The output is an individualized care plan generation model and a health needs prediction model.

[1574] Step 5: Generate an individualized care plan

[1575] The user inputs basic information (age, gender, medical history) and daily activity data through the terminal. The system receives basic information and daily activity data from the user as input. Specifically, the data is entered through an input form on the terminal and sent to the server. The output is the collected basic information and activity data.

[1576] The server generates a personalized care plan using a generative AI model based on the preprocessed health data and emotion data. It receives the preprocessed data and a specific model as input. Specifically, it generates a care plan that includes recommended activities such as "walk 7,000 steps per day" and "communicate with friends three times a week." The output is a personalized care plan.

[1577] Step 6: View Care Plan

[1578] The device displays the generated personalized care plan on the home screen or a dedicated application screen. It receives the generated care plan as input. Specifically, it visualizes the activities the user should perform that day and recommended health behaviors. The output is a visually presented care plan for the user.

[1579] Step 7: Anticipate health needs

[1580] The server analyzes current health, lifestyle, and emotional data and predicts future health risks using a generative AI model. It receives real-time data and a generative AI model as input. Specifically, it predicts the risk of nutritional deficiencies and high blood pressure. The output is health risk prediction data.

[1581] Based on the prediction results, the system suggests necessary health management actions to the user. It receives health risk prediction data as input. Specifically, it suggests taking specific supplements or undergoing regular medical checkups. The output is the suggested health management actions.

[1582] Step 8: Working with notifications

[1583] The terminal receives the prediction results and the personalized care plan from the server and sends them to the care staff's terminal. The terminal receives the prediction results and the care plan as input. Specifically, it displays a notification on the care staff's smartphone or tablet. The output is notification information for the care staff.

[1584] Based on the received information, care staff plan and implement specific care responses for the user. As input, they receive care plans and health risk prediction data. Specifically, they adjust the user's visit schedule and instruct specific health management actions. The output is the implemented care actions.

[1585] Step 9: Gather feedback

[1586] The user inputs the results of the care plan, their impressions, and changes in their physical condition via their device, and provides feedback to the server. Feedback data from the user is received as input. Specifically, feedback such as "I felt good walking today" is entered into a dedicated app. The output is the collected feedback data.

[1587] The server analyzes the collected feedback and uses it to improve and retrain the generative AI model. It receives the feedback data as input, adds new feedback, and retrains the model. The output is an improved generative AI model.

[1588] (Application example 2)

[1589] Next, a description will be given of Application Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the headset type terminal 314 will be referred to as a "terminal."

[1590] Providing an environment where elderly people can travel safely and with peace of mind is an important issue in modern society. However, conventional self-driving vehicles have difficulty providing optimal travel plans that take into account the health and emotional state of each elderly person. Furthermore, systems that can respond quickly in emergencies are inadequate. Therefore, there is a need for a system that can balance health management and transportation safety for elderly people.

[1591] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 2 is realized by the following means.

[1592] In this invention, the server includes means for collecting user health data, means for preprocessing the collected health data, means for generating an individualized care plan using a generative AI model based on the preprocessed data, means for displaying the generated care plan, means for predicting future health needs using the generative AI model, means for notifying care staff of the prediction results and the care plan, means for collecting user feedback and using it to retrain the model, means for proposing an optimal travel plan based on the user's health and emotional state in the autonomous vehicle, and means for detecting abnormal health conditions and sending emergency notifications in real time. This allows for providing an optimal travel plan that takes into account the individual health and emotional states of the elderly person, enabling rapid response in emergencies.

[1593] "Health data" refers to a user's steps, heart rate, sleep data, medical records, health checkup results, and daily life log information.

[1594] "Preprocessing" refers to the process of removing noise from collected data, correcting missing values ​​and outliers, and extracting features.

[1595] A "generative AI model" is a model trained using deep learning algorithms to predict personalized care plans and health needs based on a user's features.

[1596] "Individualized care plan" refers to a plan that provides daily activity instructions and recommends healthy behaviors based on the user's health condition and lifestyle.

[1597] "Emotional state" refers to the type of emotion (joy, sadness, anger, etc.) assessed from the user's facial expression and tone of voice.

[1598] An "optimal travel plan" refers to a plan that suggests the most comfortable and safe travel route and travel conditions for the user based on the user's health and emotional state.

[1599] "Emergency notification" refers to a system that detects abnormalities in a user's health condition in real time and promptly notifies care staff and medical institutions.

[1600] "Feedback" refers to users reporting the results of implementing a care plan and changes in their physical condition, providing data to improve the accuracy of the generative AI model based on this.

[1601] An "autonomous vehicle" is a vehicle that uses artificial intelligence to drive autonomously and reach its destination without the user having to operate it.

[1602] A specific system for implementing this invention is an application running in a self-driving vehicle that collects health and emotional data from users and provides optimal travel plans for elderly people based on that data.

[1603] Hardware and software used

[1604] Hardware:

[1605] Autonomous vehicle control unit

[1606] Smartwatches and fitness trackers worn by seniors

[1607] Cameras and microphones installed inside the vehicle

[1608] Sensor device (vehicle vital signs monitor)

[1609] software:

[1610] Generative AI models (e.g., deep learning models using TensorFlow or PyTorch)

[1611] Natural language processing technology (e.g., NLTK, spaCy)

[1612] Emotion engines (e.g. emotion recognition APIs such as Face++ and Affectiva)

[1613] What the program does

[1614] server:

[1615] The server periodically collects step counts, heart rate, and sleep data from smartwatches and fitness trackers.

[1616] It connects with medical institution databases to obtain users' medical records and health checkup results.

[1617] Emotion data is obtained from an emotion engine that uses a camera and microphone to recognize the user's face and analyze their tone of voice.

[1618] Data preprocessing:

[1619] The server removes noise from the collected data and corrects missing and outlier values.

[1620] Standardize the data and extract daily features (e.g., average daily steps, resting heart rate, total calorie intake, emotional data).

[1621] Generative AI models:

[1622] The generative AI model generates personalized care plans based on pre-processed data, while a model for predicting health needs is also built at the same time.

[1623] Emotional data is reflected in this model, enabling highly accurate predictions.

[1624] User Interface:

[1625] The generated care plan and predicted health needs are displayed on a display inside the vehicle.

[1626] Through this display, users can check their activity and health management instructions for the day.

[1627] Emergency Response:

[1628] The server monitors the user's current health data in real time.

[1629] If an abnormality is detected, the emergency notification system will be activated quickly to notify care staff and medical institutions.

[1630] Specific examples

[1631] 1. Data Collection and Analysis

[1632] Before an elderly person gets into an autonomous vehicle, their smartwatch sends their recent health data (step count, heart rate, and sleep data) to a server.

[1633] Cameras and microphones inside the vehicle collect user emotional data, which is then analyzed by an emotion engine.

[1634] 2. Generating a travel plan

[1635] Based on the collected data, the generative AI model suggests optimal travel routes for users (for example, routes that pass through parks or routes with less congestion).

[1636] The trip plan is displayed on the vehicle's display, allowing the user to view the suggested route and any important points to note.

[1637] 3. Emergency Response

[1638] Health data is monitored in real time during the journey, and if any abnormalities are detected, the vehicle will automatically stop and notify care staff or medical institutions.

[1639] Prompt Sentence Examples

[1640] Prompt: The user is currently a senior citizen. Suggest the best travel route based on his recent health and emotional data. He has low step counts and is feeling depressed, so suggest a relaxing route through the park.

[1641] This invention is a system that enables elderly people to travel safely and comfortably according to their individual health and emotional state. Real-time data monitoring and rapid emergency response provide an environment where autonomous vehicles can be used with greater peace of mind.

[1642] The flow of the specific processing in the application example 2 will be described with reference to FIG.

[1643] Step 1: Data collection

[1644] The server periodically collects step count, heart rate, and sleep data from the user's smartwatch or fitness tracker. It then connects with medical institution databases to obtain the user's medical records and health checkup results. It also uses the device's camera and microphone to analyze the user's facial recognition and tone of voice to assess their emotional state. This data is then input into the server, which outputs a comprehensive dataset on the user's health and emotional state based on the input data.

[1645] Step 2: Data Preprocessing

[1646] The server performs noise removal on the data collected in step 1 and corrects missing and outlier values. Next, it standardizes the data and extracts features such as average daily steps, resting heart rate, total calorie intake, and emotional data. It also converts text data (user feedback and medical records) into structured data using natural language processing technology. The standardized dataset is output.

[1647] Step 3: Generate an individualized care plan

[1648] The server generates a personalized care plan based on the pre-processed data using a generative AI model. The model is trained on a large dataset for elderly care. Based on the input data, the server outputs a care plan that includes daily activity instructions and recommended health behaviors.

[1649] Step 4: Reflecting emotional data

[1650] The server also incorporates emotional data into the generation of personalized care plans, providing highly accurate plans. Data obtained from the emotion engine is integrated with health data and analyzed using a generative AI model. A care plan corresponding to the emotional state based on the input data is output.

[1651] Step 5: View the care plan

[1652] The device then displays the generated care plan on the vehicle's internal display, where the user can view the day's activities and health management instructions. Based on the displayed data, the user can adjust their daily activities.

[1653] Step 6: Real-time health monitoring

[1654] As in step 1, the server monitors the user's health data in real time. If an abnormality is detected, an emergency notification system is activated immediately to notify care staff and medical institutions. This allows abnormality detection and response based on the input data.

[1655] Step 7: Optimize your travel plan

[1656] The server proposes an optimal travel plan based on the user's health and emotional state. It uses a generative AI model to calculate comfortable travel routes and travel conditions. Based on the input data, an optimized travel plan is output.

[1657] Step 8: Implement emergency response

[1658] The device detects abnormalities in the user's health condition in real time and sends an emergency notification immediately. The vehicle automatically stops and notifies pre-determined care staff and medical institutions. Emergency response is carried out based on the input data.

[1659] Step 9: Gather feedback

[1660] Users input the results of their care plan implementation and changes in their physical condition via their device. The server collects this feedback and uses it to retrain the generative AI model, improving the accuracy and effectiveness of the entire system. The retrained model is output based on the input data.

[1661] The specific processing unit 290 transmits the result of the specific processing to the headset type terminal 314. In the headset type terminal 314, the control unit 46A causes the speaker 240 and the display 343 to output the result of the specific processing. The microphone 238 acquires audio indicating a user input regarding the result of the specific processing. The control unit 46A transmits audio data indicating the user input acquired by the microphone 238 to the data processing device 12. In the data processing device 12, the specific processing unit 290 acquires the audio data.

[1662] The data generation model 58 is a so-called generative AI (Artificial Intelligence). An example of the data generation model 58 is ChatGPT (Internet Search<URL: https: / / openai.com / blog / chatgpt> ), Gemini (Internet search <url: https: gemini.google.com ?hl="ja">) and other generation AIs. The data generation model 58 is obtained by performing deep learning on a neural network. A prompt including an instruction is input to the data generation model 58, and inference data such as voice data indicating voice, text data indicating text, and image data indicating an image is also input. The data generation model 58 performs inference on the input inference data in accordance with the instruction indicated by the prompt, and outputs the inference result in a data format such as voice data and text data. Here, inference refers to, for example, analysis, classification, prediction, and / or summarization.

[1663] In the above embodiment, an example was given in which the specific processing is performed by the data processing device 12, but the technology of the present disclosure is not limited to this, and the specific processing may be performed by the headset type terminal 314.

[1664] [Fourth embodiment]

[1665] FIG. 7 shows an example of the configuration of a data processing system 410 according to the fourth embodiment.

[1666] 7, a data processing system 410 includes a data processing device 12 and a robot 414. An example of the data processing device 12 is a server.

[1667] The data processing device 12 includes a computer 22, a database 24, and a communication I / F 26. The computer 22 is an example of a "computer" according to the technology of the present disclosure. The computer 22 includes a processor 28, a RAM 30, and a storage 32. The processor 28, the RAM 30, and the storage 32 are connected to a bus 34. The database 24 and the communication I / F 26 are also connected to the bus 34. The communication I / F 26 is connected to a network 54. Examples of the network 54 include a WAN (Wide Area Network) and / or a LAN (Local Area Network).

[1668] The robot 414 includes a computer 36, a microphone 238, a speaker 240, a camera 42, a communication I / F 44, and a control target 443. The computer 36 includes a processor 46, a RAM 48, and a storage 50. The processor 46, the RAM 48, and the storage 50 are connected to a bus 52. The microphone 238, the speaker 240, the camera 42, and the control target 443 are also connected to the bus 52.

[1669] The microphone 238 receives instructions and the like from the user 20 by receiving voice uttered by the user 20. The microphone 238 captures the voice uttered by the user 20, converts the captured voice into audio data, and outputs it to the processor 46. The speaker 240 outputs audio in accordance with instructions from the processor 46.

[1670] Camera 42 is a small digital camera equipped with an optical system including a lens, aperture, and shutter, and an imaging element such as a CMOS (Complementary Metal-Oxide-Semiconductor) image sensor or a CCD (Charge Coupled Device) image sensor, and captures images of the surroundings of user 20 (for example, an imaging range defined by an angle of view equivalent to the field of vision of a typical healthy person).

[1671] The communication I / F 44 is connected to a network 54. The communication I / Fs 44 and 26 control the exchange of various information between the processor 46 and the processor 28 via the network 54. The exchange of various information between the processor 46 and the processor 28 using the communication I / Fs 44 and 26 is carried out in a secure state.

[1672] The control object 443 includes a display device, LEDs in the eyes, and motors for driving the arms, hands, and feet. The posture and gestures of the robot 414 are controlled by controlling the motors of the arms, hands, and feet. Some of the emotions of the robot 414 can be expressed by controlling these motors. In addition, the facial expressions of the robot 414 can also be expressed by controlling the light emission state of the LEDs in the eyes of the robot 414.

[1673] Fig. 8 shows an example of the main functions of the data processing device 12 and the robot 414. As shown in Fig. 8, in the data processing device 12, a specific process is performed by the processor 28. A specific process program 56 is stored in the storage 32.

[1674] The specific processing program 56 is an example of a "program" according to the technology of the present disclosure. The processor 28 reads the specific processing program 56 from the storage 32 and executes the read specific processing program 56 on the RAM 30. The specific processing is realized by the processor 28 operating as a specific processing unit 290 in accordance with the specific processing program 56 executed on the RAM 30.

[1675] The storage 32 stores a data generation model 58 and an emotion identification model 59. The data generation model 58 and the emotion identification model 59 are used by the identification processing unit 290.

[1676] In the robot 414, the processor 46 performs the reception output process. A reception output program 60 is stored in the storage 50. The processor 46 reads the reception output program 60 from the storage 50 and executes the read reception output program 60 on the RAM 48. The reception output process is realized by the processor 46 operating as the control unit 46A in accordance with the reception output program 60 executed on the RAM 48.

[1677] Next, a description will be given of the specific processing performed by the specific processing unit 290 of the data processing device 12. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1678] MODE FOR CARRYING OUT THE INVENTION

[1679] The present invention is a system for creating personalized care plans and predicting health needs using generative AI in Japan's elderly care service management system. Specific embodiments of the system are described below.

[1680] System Configuration

[1681] This system consists of users (elderly people or their caregivers), a server, and terminals (smartphones, tablets, PCs, etc.).

[1682] Program processing flow

[1683] 1. Data Collection

[1684] server

[1685] The server periodically collects step counts, heart rate, and sleep data from smartwatches and fitness trackers worn by users.

[1686] It connects with medical institution databases to obtain users' medical records and health checkup results.

[1687] It also collects daily activity logs, such as dietary habits, exercise, and medication intake, which are entered by the user via the device.

[1688] 2. Data Preprocessing

[1689] server

[1690] Noise is removed from the collected data and missing values ​​are corrected.

[1691] The total number of steps taken each day is calculated based on the step count data from the smartwatch, and the resting heart rate is calculated based on the heart rate data.

[1692] Text data (e.g., meal details) is converted into structured data using natural language processing technology.

[1693] 3. Model Building

[1694] server

[1695] A generative AI model is trained using large amounts of pre-collected elderly care and medical data, and the model uses deep learning techniques.

[1696] The model generates personalized care plans and predicts future health needs based on the user's features.

[1697] 4. Generate personalized care plans

[1698] User

[1699] Users enter basic information (age, gender, medical history) and daily activity data through the device.

[1700] The server uses this information to generate an individualized care plan using a generative AI model.

[1701] The generated care plan is displayed on the device, allowing the user to check daily activity plans and recommended health behaviors.

[1702] 5. Anticipating health needs

[1703] server

[1704] The server uses a generative AI model to predict future health risks (e.g., risk of developing nutritional deficiencies or high blood pressure) based on the user's current health and lifestyle data.

[1705] The prediction results are communicated to the user and care staff in real time.

[1706] 6. Collaboration with care staff

[1707] Terminal

[1708] The individualized care plan and health needs prediction results received from the server are sent to the care staff's terminal.

[1709] The care staff adjusts specific care responses for the user based on the information received.

[1710] 7. Feedback Loops

[1711] User

[1712] The user provides feedback to the server via their device about the results of the care plan, their impressions, and changes in their physical condition.

[1713] The server analyzes the collected feedback and uses it to retrain the generative AI model.

[1714] Specific examples

[1715] Example 1: Data collection and preprocessing

[1716] The server automatically collects nighttime sleep data and daytime step count data from the smartwatch worn by the user every day, and cleanses this data to extract useful features.

[1717] Example 2: Generating personalized care plans

[1718] When a user enters their regular health check data into the terminal, the server uses that data to generate an individualized care plan, such as "walking 8,000 steps per day" or "light aerobic exercise twice a week," and displays it to the user.

[1719] Example 3: Predicting health needs

[1720] The server analyzes data entered by the user, such as a recent loss of appetite, and evaluates the estimated risk of nutritional deficiency. The generative AI model then suggests taking a multivitamin supplement and sends an alert to care staff.

[1721] The system allows users to implement personalized care plans to maintain and improve their health, and it also allows care staff to respond more accurately, improving the quality of elderly care.

[1722] The processing flow will be explained below.

[1723] Step 1: Data collection

[1724] server

[1725] The server works in conjunction with smartwatches and fitness trackers to periodically collect the user's steps, heart rate, and sleep data.

[1726] It connects to medical institution databases via API to obtain users' past medical records and regular health checkup results.

[1727] Users input their daily activities and dietary information into the device, and this data is also sent to the server.

[1728] Step 2: Data Preprocessing

[1729] server

[1730] The server cleanses the collected data, corrects missing and outlier values, and removes inappropriate data.

[1731] Standardize the data and extract features (e.g., average daily steps, resting heart rate, total calorie intake).

[1732] Text data (meal details, user feedback) is converted into structured data using natural language processing technology.

[1733] Step 3: Model Building

[1734] server

[1735] The server trains generative AI models using large datasets for elderly care.

[1736] Deep learning algorithms are used for training to build models that generate personalized care plans from user features and predict health needs.

[1737] Evaluate the accuracy of the model and tune the parameters as needed.

[1738] Step 4: Generate an individualized care plan

[1739] User

[1740] Users enter their basic information (age, gender, medical history) and daily activity data through the device.

[1741] server

[1742] The server collects the information submitted by the user and generates a personalized care plan using a generative AI model.

[1743] The generated care plan includes daily activity instructions and recommended health behaviors (e.g., aiming for 5,000 steps per day, eating a balanced diet).

[1744] Step 5: View Care Plan

[1745] Terminal

[1746] The designated care plan is displayed on the device screen, allowing the user to check the day's activities and health management instructions.

[1747] Step 6: Anticipate health needs

[1748] server

[1749] By analyzing a user's current health and lifestyle data, a generative AI model predicts future health risks (e.g., nutritional deficiencies, risk of chronic diseases).

[1750] Based on the prediction results, the system suggests necessary health management actions to the user (e.g., taking specific supplements, undergoing regular medical checkups).

[1751] Step 7: Working with notifications

[1752] Terminal

[1753] The prediction results and individualized care plan received from the server are sent to the care staff's terminal.

[1754] Care staff

[1755] Based on the information received, the care staff plans and implements specific care responses for the user.

[1756] Step 8: Gather feedback

[1757] User

[1758] Users input the results of the care plan, their impressions, and changes in their physical condition through their terminals, and provide feedback to the server.

[1759] server

[1760] The collected feedback is analyzed and used to improve and retrain the generative AI model.

[1761] Through feedback loops, the accuracy and effectiveness of the model and the system as a whole are continually improved.

[1762] Example 1

[1763] Next, a description will be given of Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1764] To improve the quality of elderly care, it is important to accurately generate individualized care plans and predict future health risks. However, conventional systems have difficulty centrally managing and effectively analyzing a user's diverse health data. Furthermore, they lack sufficient collaboration with care staff, making it difficult to provide appropriate care. Furthermore, they lack a mechanism for appropriately incorporating user feedback and updating the model, making it difficult to respond to ever-changing health conditions.

[1765] The specific processing by the specific processing unit 290 of the data processing device 12 in the first embodiment is realized by the following means.

[1766] In this invention, the server includes: means for collecting health data from a user's activity tracking device; means for acquiring medical data in cooperation with a medical institution's database; means for collecting daily activity data entered by the user; means for performing noise reduction and missing value correction on the collected health data; means for calculating the total number of steps per day based on step count data and the resting heart rate based on heart rate data; means for converting text data into structured data using natural language processing technology; means for training a generative AI model using elderly care data and medical data; means for generating an individualized care plan using the generative AI model; means for displaying the generated care plan; means for predicting future health needs using the generative AI model; means for notifying care staff of the prediction results and the care plan; means for collecting user feedback and using it to retrain the model; means for care staff to adjust care responses to the user based on the care plan and prediction results received; and means for the user to provide feedback. This enables the generation of individualized care plans and prediction of health risks, improving the quality of elderly care. Furthermore, by appropriately incorporating user feedback and updating the model, care can be provided that is always tailored to the latest health status.

[1767] "User" refers to the elderly person or their caregiver who uses the system.

[1768] An "activity measuring device" is a device for collecting a user's physical activity data, and includes smart watches, fitness trackers, etc.

[1769] "Health data" refers to information such as a user's number of steps, heart rate, sleep data, medical records, health checkup results, dietary habits, exercise amount, and medication intake.

[1770] "Server" refers to the computer system that collects, preprocesses, analyzes, and stores user data, and trains and runs generative AI models.

[1771] "Noise reduction" refers to the process of removing unnecessary or erroneous information from collected data.

[1772] "Missing value correction" refers to the process of making predictions or inferences to fill in missing values ​​in data.

[1773] "Natural language processing technology" refers to technology for converting text data into a format that is easy for machines to understand.

[1774] "Generative AI model" refers to an AI (artificial intelligence) model that uses large amounts of data to generate personalized care plans and predict future health needs.

[1775] "Individualized care plan" refers to a care plan created based on a user's specific health condition and lifestyle habits.

[1776] "Health needs prediction" refers to the process of predicting future health risks and care needs based on a user's current data.

[1777] "Care staff" refers to professional people who provide care for older people.

[1778] "Feedback" refers to information provided by the user regarding the results of implementing the care plan, their impressions, and changes in their physical condition.

[1779] "Model retraining" refers to the process of updating a generative AI model with newly collected data and feedback to improve its accuracy.

[1780] This invention is a system for creating personalized care plans and predicting health needs using a generative AI model in an elderly care service management system in Japan. The system consists of users (elderly people or their caregivers), a server, and devices (smartphones, tablets, personal computers, etc.).

[1781] System Configuration

[1782] The system is configured as follows:

[1783] server

[1784] The server is responsible for:

[1785] 1. Periodically collect step count, heart rate, and sleep data from the user's activity tracking device (smartwatch or fitness tracker).

[1786] 2. Link with medical institution databases to obtain users' medical records and health checkup results.

[1787] 3. It also collects daily activity logs, such as dietary details, exercise volume, and medication intake, entered by the user through the device.

[1788] 4. Remove noise from the collected data and correct missing values.

[1789] 5. Calculate the total number of steps taken each day based on the step count data from the smartwatch, and calculate the resting heart rate based on the heart rate data.

[1790] 6. Convert text data (e.g., meal details) into structured data using natural language processing technology.

[1791] 7. Train generative AI models using elderly care and medical data, which use deep learning techniques.

[1792] 8. The model generates a personalized care plan and predicts future health needs based on the user's features.

[1793] Terminal

[1794] The device is used by users and care staff and has the following functions:

[1795] 1. The user enters basic information (age, gender, medical history) and daily activity data through the device.

[1796] 2. The device displays the personalized care plan generated using the generative AI model.

[1797] 3. Notify users and care staff in real time of health risk prediction results and the implementation results of care plans.

[1798] 4. Care staff will coordinate specific care responses for the user based on the information received.

[1799] User

[1800] Users provide daily activity data and feedback to the system:

[1801] 1. Users send health data to a server via their smartwatch or fitness tracker.

[1802] 2. The user enters information such as dietary habits and medication intake into the terminal application.

[1803] 3. Provide feedback on the results of the care plan, impressions, and changes in physical condition via the device application.

[1804] Specific examples

[1805] Example 1: Data collection and preprocessing

[1806] The server automatically collects nighttime sleep data and daytime step count data from the smartwatch worn by the user every day, and cleanses this data to extract useful features.

[1807] Example 2: Generating personalized care plans

[1808] When a user enters their regular health check data into the terminal, the server uses that data to generate an individualized care plan, such as "walking 8,000 steps per day" or "light aerobic exercise twice a week," and displays it to the user.

[1809] Example 3: Predicting health needs

[1810] The server analyzes data entered by the user, such as a recent loss of appetite, and assesses the estimated risk of nutritional deficiency. The generative AI model then suggests taking a multivitamin supplement and sends an alert to care staff. This system allows users to implement personalized care plans to maintain and improve their health. It also allows care staff to respond more accurately, improving the quality of elderly care.

[1811] The flow of the identification process in the first embodiment will be described with reference to FIG.

[1812] Step 1: Data collection

[1813] Server collects smartwatch data

[1814] Input: Step count, heart rate, and sleep data from the user's smartwatch

[1815] Data processing: Collect data from the smartwatch periodically and store it in a database

[1816] Output: Raw health data

[1817] Specific operation: When the user syncs the smartwatch every morning, the server receives the night's sleep data and the previous day's step count data and stores them in a database.

[1818] Acquisition of medical data by the server

[1819] Input: Medical records and health check results from medical institution databases

[1820] Data processing: Data is acquired through the medical institution's API and saved as user health information.

[1821] Output: Retrieved medical data

[1822] Specific operation: Outside of the medical institution's scheduled maintenance period, the server sends an API request to obtain the user's latest blood test results and diagnostic data.

[1823] Manual data collection by the server

[1824] Input: Dietary information, exercise amount, and medication intake status entered by the user via the device

[1825] Data processing: Collect input data in real time and save it in a database

[1826] Output: Daily activity data

[1827] Specific operation: After dinner, the user uses the terminal application to input the details of their meal and medication intake, and the data is sent to the server in real time.

[1828] Step 2: Data Preprocessing

[1829] Server-based noise removal and missing value correction

[1830] Input: Collected health data

[1831] Data processing: data denoising and missing value imputation

[1832] Output: Cleaned health data

[1833] Specific operation: The server cleanses the data obtained from the smartwatch and fills in missing daily step count data with data from the previous day.

[1834] Server recalculation and structuring of data

[1835] Input: Data after noise removal and correction

[1836] Data processing: Counting step count data, calculating resting heart rate, structuring text data

[1837] Output: Recalculated and structured data

[1838] Specific operation: The server aggregates the step count data and calculates the total number of steps taken for the day. It also analyzes the text of meal descriptions entered by the user, such as "salad and grilled chicken," and extracts information about calories and nutrients as structured data.

[1839] Step 3: Model Building

[1840] Server-based training of generative AI models

[1841] Input: Elderly care data, medical data

[1842] Data processing: Using deep learning techniques to train AI models

[1843] Output: A trained generative AI model

[1844] How it works: Generative AI models are trained using historical medical datasets and iteratively trained until error is minimized.

[1845] Step 4: Generate an individualized care plan

[1846] User enters basic information

[1847] Input: User's basic information (age, gender, medical history), daily activity data

[1848] Data processing: Enter basic information into the system and send it to the server

[1849] Output: User basic information data

[1850] Specific operation: A user logs in to the application and enters their age, gender, and medical history on the profile screen.

[1851] Server-based care plan generation

[1852] Input: User's basic information and daily activity data

[1853] Data processing: Generative AI models are used to generate personalized care plans

[1854] Output: Individualized Care Plan

[1855] Specific operation: The server calls up an AI model based on the basic information entered and generates a daily activity schedule, recommended exercises, and a meal plan.

[1856] Displaying care plans on devices

[1857] Input: Generated personalized care plan

[1858] Data processing: Push notification of care plan to device

[1859] Output: The care plan displayed to the user

[1860] What it does: The care plan is pushed to the device, and when the user opens the app, a detailed activity plan is displayed on the dashboard.

[1861] Step 5: Anticipate health needs

[1862] Server-based health risk prediction

[1863] Input: User's current health and lifestyle data

[1864] Data processing: Predicting future health risks using generative AI models

[1865] Output: Health risk prediction results

[1866] How it works: Analyzes recent dietary data to assess the user's risk of vitamin deficiency. If the deficiency persists, the AI ​​model predicts the risk and generates an alert.

[1867] Real-time notifications

[1868] Input: Health risk prediction results

[1869] Data processing: Real-time notification of prediction results

[1870] Output: Notification to user and care staff

[1871] Specific operation: A push notification is sent to the care staff's device, and an alert is displayed stating, "User A is at risk of high blood pressure."

[1872] Step 6: Collaborate with care staff

[1873] Sending care information by device

[1874] Input: Personalized care plan and health risk prediction results received from the server

[1875] Data processing: Sending information to care staff's terminal

[1876] Output: Care plan and health risk prediction information

[1877] Specific operation: Information from the server is sent to an app dedicated to care staff, who can check it in real time.

[1878] Care staff coordination

[1879] Input: Care plan and health risk prediction information

[1880] Data processing: Coordination of care responses

[1881] Output: Adjusted care plan

[1882] Specific operation: During regular visits, care staff use information obtained from the server to fine-tune the user's care plan.

[1883] Step 7: Feedback Loop

[1884] User feedback

[1885] Input: Care plan implementation results, impressions, changes in physical condition

[1886] Data processing: Input the feedback into the system and send it to the server

[1887] Output: User feedback information

[1888] Specific operation: When the user enters feedback into the app, such as "I've been walking for a while, and as a result, I feel better," that information is sent to the server.

[1889] Server data reuse

[1890] Input: Collected feedback data

[1891] Data processing: Retraining generative AI models

[1892] Output: Improved generative AI model

[1893] Specific operation: The server analyzes the collected feedback data and performs re-training to improve the accuracy of the AI ​​model.

[1894] (Application example 1)

[1895] Next, a description will be given of Application Example 1. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[1896] In conventional elderly care services, collecting health data and creating care plans is often done manually, making it difficult to provide personalized care. Furthermore, there is a lack of means to monitor health conditions in real time and quickly respond as needed. This poses a particular challenge when an elderly person becomes ill while traveling or when emergency care is required.

[1897] The specific processing by the specific processing unit 290 of the data processing device 12 in the application example 1 is realized by the following means.

[1898] In this invention, the server includes means for collecting user health data, means for preprocessing the collected health data, means for generating an individualized care plan based on the preprocessed data using a generative AI model, means for displaying the generated care plan in the vehicle, means for predicting future health needs using the generative AI model, means for notifying a caregiver of the prediction results and the care plan, means for collecting feedback from the user and using it to retrain the model, means for optimizing a driving route according to the user's health condition, and means for detecting and notifying abnormal health data in real time, thereby enabling the provision of an individualized care plan, real-time health condition monitoring, and appropriate responses even while traveling.

[1899] "User" refers to an individual about whom health data is collected, particularly the elderly and those in need of health management.

[1900] "Health data" refers to information that indicates an individual's health status, such as heart rate, number of steps, sleep data, dietary content, amount of exercise, and medication intake.

[1901] "Preprocessing" refers to the processing of collected health data, such as noise removal, missing value correction, and data structuring.

[1902] A "generative AI model" is an artificial intelligence model trained using deep learning techniques to generate personalized care plans based on collected health data and predict future health needs.

[1903] A "personalized care plan" is a specific action plan created based on the user's individual health data and lifestyle, with the aim of maintaining or improving the health of a specific user.

[1904] An "autonomous vehicle" is a vehicle that is capable of driving independently and is equipped with the functionality to collect and process health data.

[1905] "Display means" refers to a device or function for visually showing the generated care plan to the user or caregiver (e.g., an in-car display or a smartphone).

[1906] "Health needs" are the activities and care requirements necessary to maintain or improve the user's future health status.

[1907] "Care aides" are professionals and caregivers who provide care to older people and people with health care needs.

[1908] "Notification means" refers to a device or system that notifies caregivers of the generated care plan and the predicted health needs (e.g., a smartphone app or an in-car system).

[1909] "Feedback" is information that the user reports, such as the results of implementing the care plan, their impressions, and changes in their physical condition.

[1910] "Model retraining" is a machine learning process that uses collected feedback data to further improve a generative AI model.

[1911] "Route optimization" is the process of selecting the optimal driving route and rest points based on the user's health condition.

[1912] "Real-time detection" is a function that can immediately recognize and respond to abnormalities in health data when they occur.

[1913] The present invention relates to a system for providing personalized health care in autonomous vehicles, particularly designed for the elderly and people with health care needs. Specific embodiments of the system are described below.

[1914] System Configuration

[1915] This system consists of a user (such as an elderly person), a server (the autonomous vehicle's infotainment system), a device (such as a smartwatch or fitness tracker) for collecting health data, and a caregiver (such as a caregiver or medical staff).

[1916] Program processing flow

[1917] Data collection

[1918] The server collects real-time health data such as heart rate, step count, and sleep data from the smartwatch or fitness tracker worn by the user. It also connects to medical institution databases to obtain the user's medical records and health checkup results. It also collects dietary information and daily activity logs entered by the user through a terminal inside the autonomous vehicle.

[1919] Data Preprocessing

[1920] The server removes noise from the collected health data and corrects missing values. It calculates the total number of steps taken each day based on the step count data from the smartwatch and calculates the resting heart rate based on the heart rate data. Text data such as meal details is converted into structured data using natural language processing technology.

[1921] Model Building

[1922] The server uses large amounts of pre-collected elderly care and medical data to train a generative AI model, which uses deep learning techniques to generate personalized care plans and predict future health needs based on the user's features.

[1923] Personalized care plan generation

[1924] Users enter basic information (age, gender, medical history) and daily activity data through the infotainment system. The server uses this information to generate a personalized care plan using a generative AI model. The generated care plan is displayed on a display inside the autonomous vehicle, allowing users to check their daily activity plan and recommended health behaviors.

[1925] Anticipating health needs

[1926] The server uses a generative AI model to predict future health risks (e.g., risk of developing nutritional deficiencies or high blood pressure) based on the user's current health and lifestyle data. The prediction results are notified to the user and caregivers in real time.

[1927] Cooperation with care staff

[1928] The personalized care plan and health needs prediction results received from the server are sent to the caregiver's device, who then adjusts specific care responses for the user based on the received information.

[1929] Feedback Loop

[1930] Users provide feedback to the server via their devices about the results of implementing the care plan, their impressions, and changes in their physical condition. The server analyzes the collected feedback and uses it to retrain the generative AI model.

[1931] Hardware and software used

[1932] Smartwatch: Collects heart rate, steps, and sleep data.

[1933] Autonomous vehicles: Infotainment systems for collecting and processing health data.

[1934] Server: Preprocessing data, training and running generative AI models.

[1935] Deep learning framework (TensorFlow or PyTorch): Implementing and training generative AI models.

[1936] Natural language processing technologies (SpaCy, NLTK, etc.): Structuring text data.

[1937] Specific examples

[1938] For example, if a user's heart rate increases while riding in an autonomous vehicle, the autonomous vehicle's system will collect heart rate data in real time and provide a care plan to "take a deep breath and relax" and a notification recommending a break at the next service area.

[1939] Prompt Sentence Examples

[1940] "The system monitors passengers' health status in real time based on heart rate and step count data during the ride, and immediately proposes an appropriate care plan if an abnormality is detected. It also suggests optimal routes and rest points depending on the passenger's condition."

[1941] The system enables personalized care plans, real-time health monitoring and appropriate responses, even while on the move.

[1942] The flow of the specific processing in the application example 1 will be described with reference to FIG.

[1943] Step 1:

[1944] A user wears a smartwatch or fitness tracker and gets into an autonomous vehicle. The device collects real-time heart rate, step count, and sleep data from the smartwatch and sends them to a server. The input of this step is the health data from the smartwatch, and the output is the raw data sent to the server.

[1945] Step 2:

[1946] The server preprocesses the received health data. Specifically, it removes noise and corrects missing values. For example, it complements missing data based on past data and removes obviously erroneous data. The input is the raw data collected in step 1, and the output is the cleansed health data.

[1947] Step 3:

[1948] The server uses a generative AI model to generate a personalized care plan based on the preprocessed data. For example, it may recommend deep breathing if the heart rate is high or light exercise if the step count is low. The input for this step is the preprocessed health data, and the output is a personalized care plan.

[1949] Step 4:

[1950] The generated care plan is presented to the user through the autonomous vehicle's infotainment system, where the user can visually review it and follow the instructions. The input is the generated personalized care plan, and the output is the care plan displayed to the user.

[1951] Step 5:

[1952] The server uses the generative AI model to predict the user's future health needs. For example, it evaluates the risk of rising blood pressure several months from now based on current data. The input of this step is the preprocessed health data, and the output is the predicted health risks and future health needs.

[1953] Step 6:

[1954] The server notifies the caregiver of the prediction results and the care plan. The caregiver can then adjust specific care responses for the user based on this information. The input is the prediction results and the care plan, and the output is a notification sent to the caregiver.

[1955] Step 7:

[1956] The server provides an interface for collecting feedback from users. Users input the results of implementing the care plan, their impressions, and changes in their physical condition. The input is the user's feedback, and the output is the feedback data stored on the server.

[1957] Step 8:

[1958] The server analyzes the collected feedback and uses it to retrain the generative AI model, which enables the model to provide more accurate care plans. The input is the feedback data, and the output is the retrained generative AI model.

[1959] Step 9:

[1960] The server optimizes the driving route according to the user's health condition. For example, it suggests appropriate rest areas according to the user's fatigue level. The input is health data detected in real time, and the output is the optimized driving route.

[1961] Step 10:

[1962] The server detects abnormal health data in real time and immediately notifies the user and caregivers, for example, sending an alert if the heart rate is abnormally high. The input is the health data collected in real time, and the output is the alert notification sent.

[1963] Furthermore, an emotion engine that estimates the user's emotion may be further combined. That is, the identification processing unit 290 may estimate the user's emotion using the emotion identification model 59, and perform identification processing using the user's emotion.

[1964] MODE FOR CARRYING OUT THE INVENTION

[1965] The present invention is a system for creating personalized care plans and predicting health needs using generative AI and an emotion engine in an elderly care service management system in Japan. Specific embodiments of the system are described below.

[1966] System Configuration

[1967] This system consists of users (elderly people or their caregivers), a server, terminals (smartphones, tablets, PCs, etc.), and an emotion engine.

[1968] Program processing flow

[1969] 1. Data Collection

[1970] server

[1971] The server periodically collects step counts, heart rate, and sleep data from smartwatches and fitness trackers worn by users.

[1972] It connects with medical institution databases to obtain users' medical records and health checkup results.

[1973] It also collects daily activity logs, such as dietary habits, exercise, and medication intake, which are entered by the user via the device.

[1974] 2. Collecting Emotional Data

[1975] Emotion Engine

[1976] The emotion engine uses facial recognition technology to analyze the user's facial expressions and assess their emotional state (e.g., happy, sad, or angry).

[1977] Analyze the tone of the user's voice to recognize their emotional state.

[1978] 3. Data Preprocessing

[1979] server

[1980] Noise is removed from the collected health and emotion data, missing and outliers are corrected, and inappropriate data is removed.

[1981] The data is standardized and features (e.g., average daily steps, resting heart rate, total calorie intake, and emotional data) are extracted.

[1982] Text data (meal details, user feedback) is converted into structured data using natural language processing technology.

[1983] 4. Model Building

[1984] server

[1985] A generative AI model is trained using a deep learning algorithm on a large dataset for elderly care.

[1986] The model builds a model that generates an individualized care plan based on the user's features and a model that predicts health needs.

[1987] It also takes into account sentiment data, evaluates the accuracy of the model, and tunes parameters as needed.

[1988] 5. Generate personalized care plans

[1989] User

[1990] Users enter basic information (age, gender, medical history) and daily activity data through the device.

[1991] server

[1992] The server collects the emotion data from the emotion engine and the information submitted by the user, and generates a personalized care plan using a generative AI model.

[1993] The generated care plan includes daily activity instructions and recommended health behaviors (e.g., aiming for 5,000 steps per day, eating a balanced diet).

[1994] 6. Care plan display

[1995] Terminal

[1996] The care plan is displayed on the device screen, allowing users to check their activities and health management instructions for the day.

[1997] 7. Anticipating health needs

[1998] server

[1999] By analyzing a user's current health, lifestyle, and emotional data, a generative AI model predicts future health risks (e.g., risk of nutritional deficiencies and high blood pressure).

[2000] Based on the prediction results, the system suggests necessary health management actions to the user (e.g., taking specific supplements, undergoing regular medical checkups).

[2001] 8. Notifications and Collaboration

[2002] Terminal

[2003] The prediction results and individualized care plan received from the server are sent to the care staff's terminal.

[2004] Care staff

[2005] Based on the information received, the care staff plans and implements specific care responses for the user.

[2006] 9. Feedback Collection

[2007] User

[2008] Users input the results of the care plan, their impressions, and changes in their physical condition through their terminals, and provide feedback to the server.

[2009] server

[2010] The collected feedback is analyzed and used to improve and retrain the generative AI model.

[2011] Through feedback loops, the accuracy and effectiveness of the model and the system as a whole are continually improved.

[2012] Specific examples

[2013] Example 1: Data collection and preprocessing

[2014] The user collects daily step counts and heart rate data from the smartwatch, and the server cleans the data to extract useful features. At the same time, the emotion engine collects the user's emotional data using the device's camera and microphone.

[2015] Example 2: Generating personalized care plans

[2016] Users who tend to be isolated input data from regular health checkups and daily life into the device, and the server generates an individualized care plan based on that data and emotional data obtained from the emotion engine, including recommendations such as "walking 7,000 steps a day" and "communicating with friends three times a week," and displays it to the user.

[2017] Example 3: Predicting health needs

[2018] The server analyzes data showing that the user enjoys walking but has recently been feeling depressed, and the generative AI model predicts that the user has a tendency toward seasonal depression. Reflecting the data from the emotion engine, the server recommends the user engage in daily light exercise and sun exposure, and notifies care staff to contact the user regularly.

[2019] The system allows users to receive scientifically-backed, personalized care plans and responses tailored to their emotional state, helping them maintain and improve their physical and mental health, while providing caregivers with more accurate information to provide specific care.

[2020] The processing flow will be explained below.

[2021] Step 1: Data collection

[2022] server

[2023] The server works in conjunction with smartwatches and fitness trackers to periodically collect the user's steps, heart rate, and sleep data.

[2024] It connects to medical institution databases via API to obtain users' past medical records and regular health checkup results.

[2025] Daily activity logs, such as what the user eats, how much exercise they do, and what medications they take, are also sent to the server.

[2026] Step 2: Collecting emotion data

[2027] Emotion Engine

[2028] The emotion engine uses the device's camera to capture the user's facial expressions and analyzes their emotional state (happiness, sadness, anger, etc.) using facial recognition technology.

[2029] When a user speaks using the device, the tone of their voice is analyzed through a microphone to recognize their emotional state.

[2030] Step 3: Data Preprocessing

[2031] server

[2032] The server cleanses the collected health and emotion data and corrects missing or outlier values.

[2033] The data is standardized and features (e.g., average daily steps, resting heart rate, total calorie intake, and emotional data) are extracted.

[2034] Text data (meal details, user feedback) is converted into structured data using natural language processing technology.

[2035] Step 4: Model Building

[2036] server

[2037] A generative AI model is trained using a large dataset for elderly care, using deep learning algorithms.

[2038] The model will be constructed to generate an individualized care plan based on the user's features, and to predict health needs.

[2039] The accuracy of the model is evaluated taking into account sentiment data and parameters are adjusted as needed.

[2040] Step 5: Generate an individualized care plan

[2041] User

[2042] Users enter their basic information (age, gender, medical history) and daily activity data into the device.

[2043] server

[2044] The server collects the emotion data from the emotion engine and the information submitted by the user, and generates a personalized care plan using a generative AI model.

[2045] The user's plan includes daily activity instructions and recommended health behaviors (e.g., aiming for 5,000 steps per day, adjusting diet, and taking mental health measures).

[2046] Step 6: View Care Plan

[2047] Terminal

[2048] The generated care plan is displayed on the terminal screen, and the user can check the day's activities and health management instructions.

[2049] Step 7: Anticipate health needs

[2050] server

[2051] The server analyzes the user's current health, lifestyle, and emotional data, and uses a generative AI model to predict future health risks (e.g., risk of nutritional deficiencies and high blood pressure).

[2052] Based on the prediction results, the system suggests necessary health management actions to the user (e.g., taking specific supplements, undergoing regular medical checkups).

[2053] Step 8: Working with notifications

[2054] Terminal

[2055] The prediction results and individualized care plan received from the server are sent to the care staff's terminal.

[2056] Care staff

[2057] Based on the information received, the care staff plans and implements specific care responses for the user.

[2058] Step 9: Gather feedback

[2059] User

[2060] The user inputs the results of the care plan, their impressions, and any changes in their physical condition into the terminal and provides feedback to the server.

[2061] server

[2062] The collected feedback is analyzed and used to improve and retrain the generative AI model.

[2063] Through feedback loops, the accuracy and effectiveness of the model and the system as a whole are continually improved.

[2064] Example 2

[2065] Next, a description will be given of Example 2. In the following description, the data processing device 12 will be referred to as a "server" and the robot 414 will be referred to as a "terminal."

[2066] In elderly care services, it is challenging to provide care plans that take into account the user's individual health and emotional state. While conventional systems focus on collecting and analyzing health data, incorporating the user's emotional data is needed to provide more accurate personalized care plans. However, such systems face challenges due to the complexity of data collection and the need to assess the user's emotional state in real time. Furthermore, a means of clearly communicating the generated care plan and health risk predictions to the user is also required, and a mechanism for seamlessly aggregating subsequent feedback and retraining the model is lacking.

[2067] The identification processing by the identification processing unit 290 of the data processing device 12 in Example 2 is realized by the following means. In this invention, the server includes means for collecting user health data, means for preprocessing the collected health data, means for generating an individualized care plan using a generative AI model based on the preprocessed data, means for collecting user emotion data using facial recognition technology and voice analysis technology, means for preprocessing the emotion data, means for generating an individualized care plan using the preprocessed emotion data and health data, means for displaying the generated care plan, means for predicting future health needs using the generative AI model, means for notifying care staff of the prediction results and the care plan, and means for collecting feedback from the user and using it to retrain the model. This makes it possible to provide an accurate individualized care plan based on the user's health condition and emotional state, and is expected to improve the accuracy of health risk predictions and the quality of care.

[2068] A "user" is an entity that utilizes the system to provide health and emotional data and receive a personalized care plan.

[2069] "Health data" refers to data collected by users in their daily lives, such as the number of steps taken, heart rate, sleep data, dietary content, amount of exercise, and medication intake.

[2070] "Emotional data" refers to data relating to the emotional state of a user obtained from facial expressions and tone of voice collected using face recognition technology and voice analysis technology.

[2071] "Facial recognition technology" is a technology that uses a camera to analyze a user's facial expressions and assess their emotional state.

[2072] "Voice analysis technology" is a technology that uses a microphone to analyze the tone and intonation of a user's voice to assess their emotional state.

[2073] A "generative AI model" is an algorithmic model that uses collected health and emotional data to generate personalized care plans and predict future health needs.

[2074] A "personalized care plan" is a plan that includes health care instructions and activity recommendations that are optimal for a user, generated based on the user's health and emotional data.

[2075] "Preprocessing" refers to processing of collected data, such as noise removal, missing value correction, outlier removal, data standardization, and feature extraction.

[2076] "Feedback" is information provided by the user regarding the results of implementing the care plan, their impressions, and changes in their physical condition, and is used to retrain the model.

[2077] "Care staff" are people who plan and implement specific care responses for users based on the individualized care plans and health needs prediction results generated using the system.

[2078] "Notification" refers to the act of notifying the user or care staff of the generated care plan or prediction results via a terminal.

[2079] MODE FOR CARRYING OUT THE INVENTION

[2080] This invention is a system for creating personalized care plans and predicting health needs in elderly care services using generative AI and an emotion engine. This system consists of a user, a server, a terminal, and an emotion engine.

[2081] 1. Users

[2082] Users wear a smartwatch or fitness tracker to record their daily activity data. They also input their daily activity log, including their diet, exercise, and medication intake, via a device (smartphone, tablet, PC, etc.). Furthermore, the device's camera and microphone are used to record data on their emotional state.

[2083] 2. Server

[2084] The server does the following:

[2085] Data collection: The server periodically collects data from the user's smartwatch or fitness tracker and connects with medical institution databases to obtain medical records and health checkup results.

[2086] Emotional data collection: We collect user emotional data using facial recognition and voice analysis technologies.

[2087] Data preprocessing: Noise removal, missing value correction, outlier removal, standardization, feature extraction, and text data structuring are performed on the collected health and emotion data.

[2088] Model Building: Train generative AI models using large-scale elderly care datasets to build and evaluate models that generate personalized care plans and predict health needs.

[2089] Prediction and generation: Using a generative AI model, future health risks are predicted based on the user's features and emotional data, and an individualized care plan is generated.

[2090] Notification and collaboration: Prediction results and care plans are notified to care staff devices.

[2091] 3. Terminal

[2092] The terminal has the following features:

[2093] Feedback collection: Users input the results of the care plan, their impressions, and changes in their physical condition via their terminal and provide feedback to the server.

[2094] Care plan display: The generated care plan is displayed on the terminal screen, allowing the user to check the health care instructions for the day.

[2095] Notification: Display the prediction results and care plan sent from the server and notify the care staff.

[2096] Specific examples

[2097] Example 1: Data collection and preprocessing

[2098] The user records their daily steps and heart rate on their smartwatch, and the server cleans and extracts features from the data. The emotion engine collects the user's emotion data using the device's camera and microphone.

[2099] Example 2: Generating personalized care plans

[2100] Users who tend to be isolated input data from regular health checkups and daily life into the device, and the server generates an individualized care plan based on that data and emotional data obtained from the emotion engine, including recommendations such as "walking 7,000 steps a day" and "communicating with friends three times a week," and displays it to the user.

[2101] Example 3: Predicting health needs

[2102] The server analyzes data showing that the user enjoys walking but has recently been feeling depressed, and the generative AI model predicts that the user has a tendency toward seasonal depression. Reflecting the data from the emotion engine, the server recommends the user engage in daily light exercise and sun exposure, and notifies care staff to contact the user regularly.

[2103] Prompt Sentence Examples

[2104] "The user is 70 years old, their primary daily activity is walking, and the emotion engine has detected their recent emotional state as 'sad'. Please generate a personalized care plan for this user."

[2105] The flow of the identification process in the second embodiment will be described with reference to FIG.

[2106] System program processing flow

[2107] Step 1: Data collection

[2108] The server periodically collects data such as step counts, heart rate, and sleep data from the smartwatch or fitness tracker worn by the user. As input, it receives biometric data from the smartwatch or fitness tracker. Specifically, it makes API calls from the smart device and stores the data in the server's database.

[2109] The server connects to the medical institution's database to retrieve the user's medical records and health checkup results. It receives data obtained from the medical institution's API as input. Specifically, it calls the API on a regular schedule, retrieves the necessary data, and stores it in the server's database.

[2110] Users input their daily activity logs, including their diet, exercise, and medication intake, via a device. The system receives inputs from users' manual entries and log data written on a daily basis. Specifically, data is entered through a dedicated application on the device and sent to a server. The output is the collected health data and daily activity log.

[2111] Step 2: Collecting emotion data

[2112] The emotion engine uses facial recognition technology to analyze video data acquired from the device's camera and analyze the user's facial expressions. It receives the video data acquired from the camera as input. Specifically, it uses a facial recognition algorithm to analyze the facial expressions and evaluate the user's emotional state, such as "happiness" or "sadness." The output is data indicating the emotional state.

[2113] The emotion engine analyzes audio data recorded using a microphone and recognizes emotions based on the user's tone of voice. It receives audio data from the microphone as input, analyzes the audio data using a voice analysis algorithm, and evaluates the user's emotional state. The output is audio data that indicates the user's emotional state.

[2114] Step 3: Data Preprocessing

[2115] The server performs noise removal on the collected health and emotion data. As input, it receives the collected raw data. Specifically, it detects outliers and missing values ​​and removes or corrects them. The output is the cleansed data.

[2116] The server normalizes the data and extracts features such as average daily steps, resting heart rate, total calorie intake, and emotional data. It receives the cleansed data as input and applies data normalization and feature extraction algorithms to convert it into analyzable data. The output is feature data for analysis.

[2117] The server converts text data (meal details, user feedback) into structured data using natural language processing technology. It receives text data entered by the user as input. Specifically, it applies a text analysis algorithm to convert the data into structured data, such as meal items. The output is structured text data.

[2118] Step 4: Model Building

[2119] The server trains a generative AI model using a large-scale dataset for elderly care. It receives a large amount of training data as input. Specifically, it uses a deep learning algorithm (e.g., TensorFlow or PyTorch) to train the model. The output is a trained generative AI model.

[2120] The server builds a model that generates an individualized care plan and a model that predicts...

Claims

1. means for collecting health data of a user; a means for pre-processing the collected health data; a means for generating an individualized care plan using a generative AI model based on the preprocessed data; and a means for displaying the generated care plan; A means of predicting future health needs using generative AI models; and a means for notifying care staff of the predicted results and care plan; A means to collect user feedback and use it to retrain the model Including system.

2. The system of claim 1 , wherein the system receives user input and collects data on a screen device.

3. The system of claim 1 , wherein the generative AI model uses a deep learning model.

Images