Diabetes education and training platform combining virtual scene and ai video qr code

Abstract

The application discloses a diabetes education and training platform combining a virtual scene and an AI video two-dimensional code, and belongs to the technical field of digital health. The platform comprises the following modules: a user data acquisition module, which acquires multi-source data of users; a user portrait construction module, which constructs a dynamic user portrait based on data; an AI video generation module, which automatically generates a personalized education video according to the portrait and a knowledge graph; a two-dimensional code generation and distribution module, which generates a traceable dynamic two-dimensional code for the video; a mobile terminal, which is used for playing the video and providing virtual scene interaction; a virtual scene interaction module, which constructs a simulated real scene for user practice; a learning feedback and optimization module, which collects user behavior data and physiological improvement data, and continuously optimizes the user portrait and the video generation model. Through personalized content generation, scene-based practice and closed-loop optimization, the application solves the problems of traditional education content homogeneity, low participation and separation of knowledge and practice, and significantly improves the education effect and user compliance.

Description

Technical Field

[0001] This invention relates to the field of digital health technology, and in particular to a diabetes education and training platform that combines virtual scenes with AI video QR codes. Background Technology

[0002] Diabetes is a chronic disease requiring long-term self-management, making patient education and training crucial. Traditional diabetes education models primarily rely on verbal instruction from healthcare professionals, distribution of printed materials, or the showing of generic educational videos. These methods have significant shortcomings: First, the content is highly homogenized, failing to meet the personalized needs of patients with varying disease stages, cognitive levels, and lifestyles; second, the format is often monotonous, resulting in low user engagement and retention rates; and finally, the educational content is severely disconnected from patients' daily lives, leaving them with theoretical knowledge but unsure how to apply it in specific scenarios (such as supermarket shopping or dining out).

[0003] Several digital health solutions have emerged in the existing technology landscape. For example, some mobile applications (apps) offer diabetes knowledge bases and diet tracking features. Other solutions are beginning to explore simulated scenarios for patients to practice. However, these solutions are often fragmented: knowledge base apps lack in-depth personalized content and offline integration; the scenarios have fixed content and lack linkage with users' real-time physiological data and personal progress.

[0004] QR code technology has been widely used for information transmission due to its convenience. However, in the field of medical education, it is mostly used for static information redirection, such as linking to a public webpage, lacking dynamism, personalization, and deep integration with core educational processes.

[0005] Therefore, there is an urgent need in this field for an innovative solution that can deeply integrate personalized content generation, scenario training, and convenient content distribution and feedback mechanisms to systematically improve the effectiveness of diabetes education. Summary of the Invention

[0006] The purpose of this invention is to propose a diabetes education and training platform that combines virtual scenes with AI video QR codes, aiming to solve the problems of homogenized diabetes education content, low participation, and separation of knowledge and practice in existing technologies.

[0007] This invention is implemented as follows: a diabetes education and training platform combining virtual scenes and AI video QR codes, the platform comprising: The user data acquisition module is used to acquire users’ static profile data, dynamic physiological data, and cognitive behavioral data. The user profile building module is used to build dynamically updated user profiles based on multi-source user data. The AI ​​video generation module is used to automatically synthesize the personalized diabetes education video based on the user profile and the preset knowledge graph through text generation, speech synthesis and image / animation generation technologies. The QR code generation and distribution module is used to generate a unique, traceable dynamic QR code for each of the diabetes education videos and send the generated dynamic QR code to the user's mobile terminal. Mobile terminals are used to receive and display dynamic QR codes, play educational videos, and provide virtual scene interaction interfaces; The virtual scene interaction module is used to construct at least one virtual environment that simulates a real-life scenario for users to interact with and learn from. The learning feedback and optimization module is used to collect user interaction data in virtual scenarios, learning behavior data from scanning QR codes to watch educational videos in the real world, and subsequent blood glucose change data, for analysis and feedback, to continuously optimize user profiles and AI video generation models.

[0008] Furthermore, the AI ​​video generation module also includes: Knowledge graph units are used to store structured knowledge in the field of diabetes. The Natural Language Processing Unit is used to generate corresponding video scripts based on user profiles and knowledge graphs using large language models; The video generation unit is used to automatically generate personalized diabetes education videos based on the video script, combined with speech synthesis, image or animation generation technologies.

[0009] In this embodiment of the invention, the virtual environment includes a virtual supermarket, a virtual kitchen, a virtual restaurant, and a virtual sports field.

[0010] In this embodiment of the invention, the dynamic QR code has a built-in unique user identifier; In this embodiment of the invention, the multi-source user data includes the user's static profile data, dynamic physiological data, and cognitive behavioral data.

[0011] In this embodiment of the invention, the dynamic physiological data includes at least blood glucose monitoring data, dietary record data, and exercise data; the blood glucose monitoring data is connected to the user's blood glucose meter or smart device via an API interface or Bluetooth technology.

[0012] In this embodiment of the invention, the cognitive behavior data is obtained through a cognitive assessment scale embedded in the platform or user interaction behavior data in a virtual scene.

[0013] Furthermore, the user data collection module also includes External expansion modules are used to support the input or synchronization of other key physiological data via API interfaces or Bluetooth technology.

[0014] In this embodiment of the invention, the interactive behavior data in the virtual scenario interactive learning module includes, but is not limited to, selection decisions, the error rate of completing tasks, operation sequences, task completion time, and points of focus.

[0015] In this embodiment of the invention, the learning behavior data of scanning QR codes to watch educational videos in the real world includes, but is not limited to, video viewing completion rate and viewing frequency.

[0016] Beneficial effects of the present invention This invention proposes a diabetes education and training platform combining virtual scenarios and AI video QR codes, belonging to the field of digital health technology. The platform includes: a user data collection module for acquiring static user profiles, dynamic physiological and cognitive behavioral data; a user profile construction module for building dynamic user profiles based on multi-source data; an AI video generation module for automatically synthesizing personalized educational videos using text generation, speech synthesis, and image animation technologies based on user profiles and a diabetes knowledge graph; a QR code generation and distribution module for generating a unique and traceable dynamic QR code for each video and pushing it to the user's terminal; a mobile terminal for receiving and displaying the dynamic QR code, playing educational videos, and providing a virtual scenario interaction interface; a virtual scenario interaction module for constructing a virtual environment simulating real-life scenarios for user practice; and a learning feedback and optimization module for collecting user interaction behavior and subsequent physiological improvement data after watching videos, continuously optimizing the user profile and video generation model. This invention effectively solves the problems of homogenized content, low participation, and separation of knowledge and action in traditional diabetes education through personalized content generation, scenario-based practical training, and a closed-loop feedback mechanism, significantly improving educational effectiveness and user compliance. Attached Figure Description

[0017] Figure 1 This is a structural diagram of a diabetes education and training platform that combines virtual scenes and AI video QR codes according to a preferred embodiment of the present invention. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. For ease of explanation, only the parts related to the embodiments of this invention are shown. It should be understood that the specific embodiments described herein are merely for explaining this invention and are not intended to limit this invention.

[0019] This invention proposes a diabetes education and training platform combining virtual scenarios and AI video QR codes, belonging to the field of digital health technology. The platform includes: a user data collection module for acquiring static user profiles, dynamic physiological and cognitive behavioral data; a user profile construction module for building dynamic user profiles based on multi-source data; an AI video generation module for automatically synthesizing personalized educational videos using text generation, speech synthesis, and image animation technologies based on user profiles and a diabetes knowledge graph; a QR code generation and distribution module for generating a unique and traceable dynamic QR code for each video and pushing it to the user's terminal; a mobile terminal for receiving and displaying the dynamic QR code, playing educational videos, and providing a virtual scenario interaction interface; a virtual scenario interaction module for constructing a virtual environment simulating real-life scenarios for user practice; and a learning feedback and optimization module for collecting user interaction behavior and subsequent physiological improvement data after watching videos, continuously optimizing the user profile and video generation model. This invention effectively solves the problems of homogenized content, low participation, and separation of knowledge and action in traditional diabetes education through personalized content generation, scenario-based practical training, and a closed-loop feedback mechanism, significantly improving educational effectiveness and user compliance.

[0020] Figure 1 This is a structural diagram of a diabetes education and training platform combining virtual scenes and AI video QR codes, provided by a preferred embodiment of the present invention. The platform includes: The user data acquisition module is used to acquire users’ static profile data, dynamic physiological data, and cognitive behavioral data. The user profile building module is used to build dynamically updated user profiles based on multi-source user data; the multi-source user data includes users’ static profile data, dynamic physiological data and cognitive behavior data. The AI ​​video generation module is used to automatically synthesize the personalized diabetes education video based on the user profile and the preset knowledge graph through text generation, speech synthesis and image / animation generation technologies. The QR code generation and distribution module is used to generate a unique, traceable dynamic QR code for each of the diabetes education videos and send the generated dynamic QR code to the user's mobile terminal. Mobile terminals are used to receive and display dynamic QR codes, play educational videos, and provide virtual scene interaction interfaces; The virtual scene interaction module is used to construct at least one virtual environment that simulates a real-life scenario for users to interact with and learn from. The learning feedback and optimization module is used to collect user interaction data in virtual scenarios, learning behavior data from scanning QR codes to watch educational videos in the real world, and subsequent blood glucose change data, for analysis and feedback, to continuously optimize user profiles and AI video generation models.

[0021] Furthermore, the AI ​​video generation module also includes: The knowledge graph unit is used to store structured knowledge in the field of diabetes; the knowledge in the field of diabetes includes medical knowledge, nutritional knowledge, and knowledge of complications, etc. The Natural Language Processing Unit is used to generate corresponding video scripts based on user profiles and knowledge graphs using large language models; The video generation unit is used to automatically synthesize personalized diabetes education videos based on the video script, combined with speech synthesis, image or animation generation technologies; The speech synthesis technology can use a TTS engine to convert scripts into speech narration, and the image or animation generation uses a visual engine to call a material library to generate dynamic videos. In this embodiment of the invention, the static profile data includes basic personal information (including age, height, weight, etc.), diabetes type, disease course, medication information (including drug name, dosage, etc.), etc.; it can be obtained through structured form input. The dynamic physiological data includes at least blood glucose monitoring data, dietary record data, and exercise data; The blood glucose monitoring data can be connected to the user's blood glucose meter or smart device (such as a smart bracelet) via API interface or Bluetooth technology to synchronize blood glucose values, blood glucose fluctuation trends and other data in real time or periodically. The dietary record data can be obtained by the user manually entering it into the App. The exercise data can be obtained by the user manually inputting it in the App, or by connecting to the user's smart device (such as a fitness tracker) through an API interface or Bluetooth technology. The cognitive behavior data is obtained through the cognitive assessment scale embedded in the platform or through user interaction behavior data in virtual scenarios. In one embodiment of the present invention, the user data collection module further includes External expansion modules are used to support the input or synchronization of other key physiological data, such as blood pressure, blood lipids, weight, and BMI, via API interface or Bluetooth technology, to build a more comprehensive health view.

[0022] In this embodiment of the invention, the virtual environment includes virtual supermarkets, virtual kitchens, virtual restaurants, and virtual sports fields, etc. Users can complete tasks such as food selection, meal preparation, dining out decisions, and exercise plan execution within this virtual environment. Users interact with elements in the scene to complete set health education tasks. During user interaction, the virtual scenario interactive learning module records all their interactive behavior data, including but not limited to: selection decisions (such as what food was taken), operation sequences (such as whether to test blood sugar first or inject insulin first), task completion time, and focus points (such as whether food labels were viewed). This interactive behavior data is a key input for triggering personalized content and optimizing user profiles.

[0023] The video content associated with the dynamic QR code can change dynamically according to the update of user data, and each QR code has a unique user identifier; the QR code can be distributed to users via App messages, SMS, email or printed personalized guide manuals.

[0024] The learning behavior data in the real world for scanning QR codes to watch educational videos includes, but is not limited to, video completion rate and viewing frequency.

[0025] The following example illustrates in detail the working process of the diabetes education and training platform that combines virtual scenes and AI video QR codes according to an embodiment of the present invention. User background: Mr. Zhang, 50 years old, is a newly diagnosed type 2 diabetes patient. The platform learned through the initial assessment that he usually eats white porridge, fried dough sticks and steamed buns for breakfast, and has a vague understanding of the concept of glycemic index (GI) of food.

[0026] Step 1: Data collection and user profile building; The data acquisition module collected the following multi-source data: Static record data: age, gender, disease course (newly diagnosed), medication (oral metformin); Dynamic physiological data: For three consecutive days, the blood glucose level 2 hours after breakfast was between 11.5-13.0 mmol / L, which was significantly exceeded; Cognitive behavioral data: Through the app's food log, it was found that he often recorded "one bowl of white porridge" and "one fried dough stick"; through a short cognitive test, it was found that his error rate in classifying "high GI foods" and "low GI foods" was as high as 70%; Based on the above data, the user profile building module dynamically updates Mr. Zhang's profile and adds the following tags: High risk: High blood sugar after breakfast; Behavior: Preference for high-GI foods (white porridge, refined flour) as breakfast staples; Knowledge gap: The concept and application of food GI values; Learning stage: Newly diagnosed patients require basic dietary education; Step 2: AI generates personalized educational videos and QR codes; The AI ​​video generation module is triggered, and its workflow is as follows: After receiving tags from a user profile, the knowledge graph unit automatically retrieves and associates relevant knowledge nodes: White porridge has a high glycemic index (GI); fried dough sticks are classified as high-fat and high-sugar foods; oatmeal has a low GI; whole wheat bread has a low GI; key points: "The relationship between GI value and blood sugar fluctuations", "Principles of healthy breakfast pairings"; The video generation unit begins synthesizing the video: Text generation: Based on user profiles and knowledge graphs, a large language model generates a targeted video script. The script might begin with: "Mr. Zhang, your post-breakfast blood sugar has been a bit high lately, which is likely related to your usual white porridge and fried dough sticks..." The script intersperses comparisons of stable blood sugar curves from low-GI breakfasts (such as oat milk) and offers specific suggestions for "replacing staple foods".

[0027] Speech synthesis: The TTS engine converts the script into speech narration.

[0028] Image / Animation Generation: The visual engine calls upon the material library to generate dynamic comparison animations of "blood sugar roller coaster" and "blood sugar stable walking," and displays intuitive comparison images of white porridge, fried dough sticks, oatmeal, and whole wheat bread.

[0029] Video compositing: Combine all elements into a 2-minute short video titled "Say Goodbye to the Blood Sugar Rollercoaster: Your Healthy Breakfast Guide".

[0030] The QR code generation and distribution module then generates a unique dynamic QR code for the video and pushes it to Mr. Zhang's mobile phone via App message.

[0031] Step 3: User Learning and Virtual Scene Interaction Scan to learn: The next morning, Mr. Zhang scanned the QR code and watched the entire video customized for him on his mobile phone.

[0032] Virtual scene training: After the video finishes playing, a prompt to enter the "virtual breakfast shop" automatically pops up on the screen; Mr. Zhang clicks to enter the "virtual breakfast shop" built by the virtual scene interaction module, and the system issues a task: "Please prepare a healthy breakfast for Mr. Zhang that can stabilize his post-meal blood sugar."

[0033] Mr. Zhang first tried selecting "plain congee + fried dough sticks." The system immediately simulated an animation of a post-meal blood sugar spike and displayed a prompt: "This combination will cause a rapid rise in blood sugar. Please choose again." Next, Mr. Zhang selected "oatmeal cup + milk + a boiled egg." The system immediately gave positive feedback: "Great! This is a low-GI, high-nutrient option. Your post-meal blood sugar is expected to be very stable." It also displayed a smooth simulated blood sugar curve.

[0034] Step 4: Learning Feedback and Model Optimization; The learning feedback and assessment module has begun collecting data on Mr. Zhang's interactive behavior during this learning and training session: Video learning data: 100% completion rate, watched twice.

[0035] Virtual scene data: The first selection was incorrect, the second selection was correct, and the task was completed.

[0036] Subsequent physiological data: Over the next week, the platform continued to monitor Mr. Zhang's post-breakfast blood glucose levels. Data showed that the average level decreased from 12.5 mmol / L to 8.2 mmol / L.

[0037] Analysis and optimization: The learning feedback and assessment module comprehensively analyzed the data and concluded that the educational intervention centered on the "GI concept" and "breakfast substitution" was effective.

[0038] Optimize user profile: Update Mr. Zhang's tag from "Knowledge Blind Spot: Food GI Value Concept and Application" to "Mastered: Basic Application of GI Concept"; and add a new tag "Behavioral Improvement: Starting to Try Low-GI Breakfast". At the same time, the weight of "High Risk: High Blood Sugar After Breakfast" has been reduced.

[0039] Optimize the AI ​​video generation model: This successful case will serve as a positive example to strengthen the effectiveness of the model in prioritizing the "GI concept education" and "staple food replacement" strategies when dealing with the "post-breakfast hyperglycemia" problem.

[0040] Closed loop formation: Based on the updated user profile, the platform may generate a new video and QR code for Mr. Zhang about "lunch protein intake" or "healthy snack options" in the next instance, thus initiating a new, spiraling cycle of personalized education.

[0041] The embodiments of this invention clearly demonstrate how the platform seamlessly integrates data-driven approaches, personalized content, contextualized training, and closed-loop feedback to form an intelligent and efficient digital health intervention system.

[0042] Those skilled in the art will understand that all or part of the steps in the methods of the above embodiments can be implemented by program instructions and related hardware. The program can be stored in a computer-readable storage medium, such as ROM, RAM, disk, optical disk, etc.

[0043] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A diabetes education and training platform combining virtual scenes and AI video QR codes, characterized in that, The platform includes: The user data acquisition module is used to acquire users’ static profile data, dynamic physiological data, and cognitive behavioral data. The user profile building module is used to build dynamically updated user profiles based on multi-source user data. The AI ​​video generation module is used to automatically synthesize the personalized diabetes education video based on the user profile and the preset knowledge graph through text generation, speech synthesis and image / animation generation technologies. The QR code generation and distribution module is used to generate a unique, traceable dynamic QR code for each of the diabetes education videos and send the generated dynamic QR code to the user's mobile terminal. Mobile terminals are used to receive and display dynamic QR codes, play educational videos, and provide virtual scene interaction interfaces; The virtual scene interaction module is used to construct at least one virtual environment that simulates a real-life scenario for users to interact with and learn from. The learning feedback and optimization module is used to collect user interaction data in virtual scenarios, learning behavior data from scanning QR codes to watch educational videos in the real world, and subsequent blood glucose change data, for analysis and feedback, to continuously optimize user profiles and AI video generation models.

2. The diabetes education and training platform combining virtual scenes and AI video QR codes as described in claim 1, characterized in that, The AI ​​video generation module also includes: Knowledge graph units are used to store structured knowledge in the field of diabetes. The Natural Language Processing Unit is used to generate corresponding video scripts based on user profiles and knowledge graphs using large language models; The video generation unit is used to automatically generate personalized diabetes education videos based on the video script, combined with speech synthesis, image or animation generation technologies.

3. The diabetes education and training platform combining virtual scenes and AI video QR codes as described in claim 1, characterized in that, The virtual environment includes a virtual supermarket, a virtual kitchen, a virtual restaurant, and a virtual sports field.

4. The diabetes education and training platform combining virtual scenes and AI video QR codes as described in claim 1, characterized in that, The dynamic QR code contains a unique user identifier.

5. The diabetes education and training platform combining virtual scenes and AI video QR codes as described in claim 1, characterized in that, The multi-source user data includes users' static profile data, dynamic physiological data, and cognitive behavioral data.

6. The diabetes education and training platform combining virtual scenes and AI video QR codes as described in claim 1, characterized in that, The dynamic physiological data includes at least blood glucose monitoring data, dietary record data, and exercise data; The blood glucose monitoring data is connected to the user's blood glucose meter or smart device via API interface or Bluetooth technology.

7. The diabetes education and training platform combining virtual scenes and AI video QR codes as described in claim 1, characterized in that, The cognitive behavior data is obtained through the cognitive assessment scale embedded in the platform or through user interaction behavior data in virtual scenarios.

8. The diabetes education and training platform combining virtual scenes and AI video QR codes as described in claim 1, characterized in that, The user data collection module also includes External expansion modules are used to support the input or synchronization of other key physiological data via API interfaces or Bluetooth technology.

9. The diabetes education and training platform combining virtual scenes and AI video QR codes as described in claim 1, characterized in that, The interactive behavior data in the virtual scenario interactive learning module includes, but is not limited to, selection decisions, the error rate of completing tasks, operation sequences, task completion time, and focus points.

10. The diabetes education and training platform combining virtual scenes and AI video QR codes according to claim 1, characterized in that, The learning behavior data in the real world for scanning QR codes to watch educational videos includes, but is not limited to, video completion rate and viewing frequency.

Images