A patient education system and method based on augmented reality scene simulation

Abstract

This invention discloses a patient education system and method based on augmented reality (AR) scenario simulation. The system includes an AR wearable terminal, a motion capture and data acquisition device, a virtual scene construction and interaction device, an intelligent assessment and feedback device, a cloud data management device, and a medical staff terminal. Medical staff input patient information and push suitable scenarios through the medical staff terminal. Patients wear the AR wearable terminal to enter the virtual scene. The motion capture and data acquisition device collects movement trajectories and electromyographic signal data, which are calibrated and fused for scene branch evolution and movement deviation analysis, generating an assessment report and pushing personalized supplementary educational content. This invention achieves immersive and precise patient education, improves patient participation and educational effectiveness, and reduces the burden on medical staff.

Description

Technical Field

[0001] This invention belongs to the field of patient education technology, and relates to a patient education system and method based on augmented reality scenario simulation. Background Technology

[0002] In clinical rehabilitation and patient education, traditional methods rely heavily on oral explanations, written materials, or video presentations, resulting in a lack of interactivity and a monotonous approach. Patients passively receive information, making it difficult for them to intuitively understand the dangers of risky scenarios and the standard techniques for rehabilitation movements. This leads to poor retention of the educational content and makes it challenging to effectively avoid risks or perform rehabilitation movements correctly in practice.

[0003] In traditional rehabilitation training, medical staff need to provide one-on-one guidance to patients on movement exercises, judging whether the patient's movements are standard through visual observation. This method is limited by the experience and energy of medical staff, making it difficult to accurately capture subtle deviations in movement or quantify the degree of deviation. At the same time, data reflecting muscle exertion, such as electromyography signals, cannot be effectively collected, resulting in highly subjective assessment results and making it difficult to provide targeted corrective guidance.

[0004] Patients vary in age, disease type, surgical procedure, and rehabilitation stage, resulting in different needs for educational content and rehabilitation training. Traditional educational models often employ uniform content and standards, lacking personalized adaptation and failing to meet the specific needs of different patients. Some patients may find the educational content too simplistic or complex, hindering the desired educational effect and impacting their rehabilitation progress.

[0005] Furthermore, patient information, training data, and assessment results generated during traditional patient education processes are mostly stored in paper-based or scattered formats, resulting in chaotic data management and difficulty in traceability. Healthcare professionals cannot quickly summarize and analyze patient education data, nor can they identify common problems among different patient groups. This leads to a lack of data support for optimizing education programs, hindering the improvement of overall education quality and efficiency.

[0006] Meanwhile, traditional education methods require medical staff to invest a lot of time and energy in explanation, guidance and assessment. When there are a large number of clinical patients, the workload of medical staff is heavy, making it difficult to provide detailed education and training guidance for each patient, which further affects the effectiveness of education and the standardization of rehabilitation training. Summary of the Invention

[0007] To address the problems existing in the background technology, this invention proposes a patient education system and method based on augmented reality scene simulation.

[0008] The first aspect of this application provides a patient education method based on augmented reality scenario simulation, including: Medical staff enter patients’ basic information through medical terminals, select corresponding risk simulation scenarios and rehabilitation training scenarios from the scenario library of the cloud data management device, and push them to the augmented reality wearable terminal. The patient wears an augmented reality wearable terminal to enter a virtual scene. The motion capture and data acquisition device collects the patient's motion trajectory data and electromyographic signal data in real time, which are then transmitted to the intelligent assessment and feedback device and the virtual scene construction and interaction device, respectively. The virtual scene construction and interaction device drives the branch evolution of the virtual scene based on the received motion data; the intelligent assessment and feedback device calculates the motion deviation degree through the motion deviation analysis model; and the real-time feedback unit provides voice prompts and visual prompts to the patient through the augmented reality wearable terminal. After the education campaign concludes, the education effectiveness evaluation model generates an evaluation report and uploads it to the cloud data management device. Medical staff can view the evaluation report through their medical terminals and push personalized supplementary education content to augmented reality wearable terminals.

[0009] Optionally, the intelligent assessment and feedback device achieves data fusion through a spatiotemporal synchronization calibration algorithm, including: using the action trigger time of the inertial measurement unit as the reference timestamp T0; detecting the signal trigger time T1 of the electromyography (EMG) sensor and calculating the time difference ΔT = T1 - T0; shifting the EMG signal along the time axis according to the difference between ΔT and ΔT0 to achieve time alignment between the EMG signal and the action trajectory data; and performing spatiotemporal correlation fusion between the aligned EMG signal and the action trajectory data; where ΔT0 is the EMG-action synchronization time difference for a standard action. A second aspect of this application provides a patient education system based on augmented reality scene simulation, including: an augmented reality wearable terminal, a motion capture and data acquisition device, a virtual scene construction and interaction device, an intelligent assessment and feedback device, a cloud data management device, and a medical terminal; each device interacts with data through a wireless communication protocol. Augmented reality wearable devices are used to present virtual scenes and receive voice commands from patients; The motion capture and data acquisition device is used to collect patient movement trajectory data and electromyographic signal data and transmit them to the intelligent assessment and feedback device and the virtual scene construction and interaction device. The virtual scene construction and interaction device is used to construct risk simulation sub-scenes and rehabilitation training sub-scenes, and drives the evolution of scene branches based on the received patient action data; The intelligent assessment and feedback device is used to analyze movement deviations and generate assessment results, providing feedback to patients through an augmented reality wearable terminal; The cloud-based data management device is used to store data and update the scene and parameter libraries; the medical terminal is used to view the evaluation results and push personalized educational content to the augmented reality wearable terminal.

[0010] Optionally, the augmented reality wearable terminal includes a high-definition display screen, a voice interaction unit, a positioning unit, a haptic feedback unit, and an environmental sound effect simulation unit; the high-definition display screen is used to present virtual scenes and virtual models for rehabilitation training; the voice interaction unit is used to receive voice commands from the patient; the positioning unit is used to obtain the patient's physical spatial location information; the haptic feedback unit is used to generate vibration feedback at the corresponding limbs of the patient; and the environmental sound effect simulation unit is used to play environmental sound effects corresponding to the virtual scene.

[0011] Optionally, the motion capture and data acquisition device includes an inertial measurement unit, an electromyography (EMG) sensor, and a depth camera; the inertial measurement unit is used to acquire angular velocity and acceleration data of the patient's head posture and limb movements; the EMG sensor is used to acquire the EMG signal intensity of the target muscle groups in rehabilitation training; and the depth camera is used to capture the contours of the patient's whole-body movements and generate a three-dimensional motion model.

[0012] Optionally, the virtual scene construction and interaction device includes a scene library and an interaction logic engine; the scene library contains risk simulation sub-scenes and rehabilitation training sub-scenes, the risk simulation sub-scenes preset correct behavior interaction branches and incorrect behavior interaction branches, and the rehabilitation training sub-scenes preset standard action parameter thresholds; the interaction logic engine is used to receive action data transmitted by the motion capture and data acquisition device, determine whether the patient's actions conform to the preset logic of the scene, and drive the evolution of virtual scene branches.

[0013] Optionally, the intelligent assessment and feedback device includes a movement deviation analysis model, a health education effectiveness evaluation model, and a real-time feedback unit. The movement deviation analysis model receives three-dimensional movement data, electromyographic signal data, and standard movement parameter library data from the patient, and calculates the movement deviation degree using a movement deviation degree calculation formula. The health education effectiveness evaluation model combines the accuracy rate of scene interaction, the average value of movement deviation degree, and the accuracy rate of health education questions and answers to generate the patient's mastery level. The real-time feedback unit provides movement correction information to the patient through voice prompts and visual prompts.

[0014] Optionally, the formula for calculating the motion deviation is D=α×(1-cosθ)+β×(1-S p / S s ); where D is the movement deviation; α is the movement trajectory weighting coefficient; θ is the angle between the patient's movement trajectory vector and the standard movement trajectory vector; β is the electromyographic signal weighting coefficient; S p S represents the average actual electromyographic signal intensity of the patient's muscle groups. s This represents the average electromyographic signal intensity of the muscle group corresponding to the standard movement.

[0015] Optionally, the motion deviation analysis model is configured with a dynamic weight coefficient adjustment function to adjust the values ​​of α and β based on the type of rehabilitation training; when the type of rehabilitation training is joint mobility training, α is 0.8 and β is 0.2; when the type of rehabilitation training is muscle strength training, α is 0.3 and β is 0.7; the weight coefficients can be manually adjusted through the medical terminal.

[0016] Optionally, the cloud-based data management device is used to store basic patient information, motion collection data, and education effectiveness evaluation results, build and update a risk scenario database and a rehabilitation training motion parameter database, and use big data analysis to identify weaknesses in education for different patient groups.

[0017] Compared with the prior art, the present invention has the following beneficial effects: This invention provides a patient education system and method based on augmented reality scenario simulation. In terms of the educational experience, the augmented reality wearable terminal constructs an immersive environment through multi-unit collaboration. The dynamic branching evolution of the virtual scene allows the patient's actions to directly determine the outcome. Multi-sensory feedback allows patients to intuitively experience the consequences of risks and the differences in actions, breaking the limitations of traditional one-way instructional education, significantly improving patient participation and immersion, and helping patients to more deeply remember risk avoidance methods and standard rehabilitation movements.

[0018] In terms of accuracy, motion capture and data acquisition devices enable comprehensive collection of multi-dimensional data, while spatiotemporal synchronization calibration algorithms eliminate errors caused by data asynchrony. The formula for calculating movement deviation precisely quantifies movement deviations in different types of rehabilitation training through dynamic weight adjustments, and the education effectiveness evaluation model integrates multiple indicators to form a comprehensive evaluation result. The entire evaluation process is based on objective data, avoiding the subjectivity and ambiguity of traditional qualitative assessments, making the evaluation results more scientific and convincing.

[0019] In terms of personalized adaptation, the system first precisely matches patients' basic information with a scenario database to push appropriate initial educational content to patients of different ages, diseases, and rehabilitation stages. After the educational phase, the system automatically pushes personalized supplementary content based on the assessment results, focusing on strengthening the patient's weaknesses to achieve individualized instruction. In addition, the system also provides a manual adjustment function for dynamic weight coefficients, allowing medical staff to fine-tune the system according to the patient's individual physical condition, thereby making the assessment and guidance more closely aligned with the patient's actual needs.

[0020] In terms of improving healthcare efficiency, cloud-based data management devices enable automated storage and integration of various data, generating standardized assessment reports that allow healthcare professionals to quickly grasp patient education progress. Big data analysis identifies weaknesses in patient groups, providing data support for healthcare professionals to optimize overall education programs. Automated data collection, evaluation, and content delivery reduce manual operations and repetitive tasks for healthcare professionals, lowering their workload and allowing them to focus more on clinical treatment and rehabilitation guidance for patients.

[0021] In terms of system scalability, the cloud-based data management device supports continuous updates to the risk scenario library and rehabilitation training movement parameter library, enabling the integration of new clinical cases and medical research findings. The system's modular design allows for flexible upgrades of each device's functionality, adapting to the educational needs of different clinical departments and the evolving trends in rehabilitation medicine. This scalability ensures the system can serve clinical educational work for a long time, continuously delivering value.

[0022] In summary, this invention constructs an immersive, precise, and personalized education system through the application of multi-device collaboration and intelligent technology. This system not only enhances the patient's education experience and understanding, but also optimizes the workflow and efficiency of medical staff. Furthermore, it possesses good scalability and adaptability, providing a scientific and efficient solution for clinical rehabilitation education. Attached Figure Description

[0023] Figure 1 This is a flowchart of a patient education method based on augmented reality scene simulation in one embodiment of the present invention; Figure 2 This is a schematic diagram of a patient education system based on augmented reality scene simulation in one embodiment of the present invention. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] In one embodiment, such as Figure 1 As shown, a patient education method based on augmented reality scenario simulation is provided, which can be applied to... Figure 1 Taking China as an example, the following specific steps will be used: S10: Medical staff enter patients' basic information through medical terminals, select corresponding risk simulation scenarios and rehabilitation training scenarios from the scenario library of the cloud data management device, and push them to the augmented reality wearable terminal.

[0026] Specifically, before conducting patient education, medical staff need to enter basic patient information through a medical terminal. The terminal's interface provides standardized information entry options, allowing medical staff to sequentially enter information such as the patient's age, disease type, surgical procedure, and rehabilitation training stage. This information is the core basis for subsequent scenario matching, ensuring that the selected scenario closely aligns with the patient's actual clinical needs.

[0027] After information entry is completed, the medical terminal automatically establishes data interaction with the cloud-based data management device. The cloud-based data management device's scenario library stores various preset risk simulation scenarios and rehabilitation training scenarios. Each scenario in the scenario library is associated with specific applicable conditions, which correspond to various dimensions of the patient's basic information. Medical staff can select suitable risk simulation scenarios and rehabilitation training scenarios from the scenario list displayed on the medical terminal based on the patient's specific situation. During the selection process, the medical terminal provides scenario recommendations based on the entered patient information, assisting medical staff in quickly identifying appropriate scenarios.

[0028] After identifying the target scenario, medical staff issue a push notification via their medical terminal. This notification is transmitted to a cloud data management device via a wireless communication protocol. Upon receiving the notification, the cloud data management device retrieves the corresponding scenario data and sends it to the augmented reality wearable terminal worn by the patient. After receiving the scenario data, the augmented reality wearable terminal automatically loads the scenario, ensuring that the patient can promptly enter the corresponding virtual scenario for educational experience.

[0029] Taking an elderly patient requiring knee rehabilitation after orthopedic surgery as an example, medical staff input information such as the patient's age (72 years old), condition (femoral neck fracture), surgery type (total hip replacement), and rehabilitation stage (two weeks post-surgery) into a medical terminal. Then, they select a fall risk simulation scenario and a knee flexion-extension rehabilitation training scenario from the scenario library. After issuing a push command, the augmented reality wearable terminal quickly loads these two scenarios. After wearing the terminal, the patient can first experience the fall risk scenario and then undergo targeted rehabilitation training. The entire process requires no additional operation to achieve accurate scenario delivery.

[0030] This process enables personalized and precise scenario selection, avoiding the mismatch between generic scenarios and patients' actual needs. Through the collaborative work of medical staff terminals and cloud-based data management devices, the operational procedures for medical personnel are simplified, reducing subjective errors in scenario selection. Simultaneously, rapid scenario delivery ensures efficient educational outreach, allowing patients to receive appropriate educational content at the right time, laying the foundation for subsequent immersive experiences and accurate assessments.

[0031] S20: The patient wears an augmented reality wearable terminal to enter a virtual scene. The motion capture and data acquisition device collects the patient's motion trajectory data and electromyographic signal data in real time, and transmits them to the intelligent assessment and feedback device and the virtual scene construction and interaction device, respectively.

[0032] Specifically, after receiving the notification to begin the educational session, the patient puts on the augmented reality wearable terminal. The positioning unit of the augmented reality wearable terminal automatically activates, acquiring the patient's location information in the physical space in real time to ensure that the virtual scene maintains spatial matching with the real environment and avoids scene misalignment affecting the experience. The patient views the loaded virtual scene through the high-definition display screen of the augmented reality wearable terminal, and the voice interaction unit is ready to receive the patient's voice commands at any time, allowing the patient to adjust the scene playback status during the experience.

[0033] Upon entering the virtual environment, the motion capture and data acquisition devices activate simultaneously. The inertial measurement unit (IMU) collects real-time data on the patient's head posture and limb movements, including angular velocity and acceleration, accurately recording the trajectory changes of each movement within the virtual environment. Electromyography (EMG) sensors are attached to the target muscle groups corresponding to the patient's rehabilitation training, continuously collecting the intensity of muscle electrical signals to reflect the force exerted by the patient's movements. A depth camera captures the contours of the patient's entire body, transforming two-dimensional movements into a three-dimensional motion model, achieving a comprehensive record of the patient's overall movements.

[0034] Motion trajectory data and electromyographic signal data collected by the motion capture and data acquisition device are transmitted in real time via a wireless communication protocol. A portion of the data is transmitted to the intelligent evaluation and feedback device, providing raw data support for subsequent motion deviation analysis and effect evaluation. Another portion is transmitted to the virtual scene construction and interaction device, providing a basis for judging the branching evolution of the virtual scene. The entire acquisition process is performed in real time, with data transmission latency controlled to the millisecond level, ensuring that data can be fed back to the corresponding devices in a timely manner and guaranteeing smooth interaction.

[0035] Taking stroke patients undergoing upper limb rehabilitation training as an example, patients wear augmented reality wearable terminals and enter a virtual scenario for upper limb joint flexion and extension rehabilitation training. Following the virtual model in the scenario, patients perform movements such as raising, extending, and bending their upper limbs. An inertial measurement unit collects angular velocity and acceleration data of the upper limb movements, an electromyography (EMG) sensor collects the EMG signal intensity of the biceps and triceps brachii muscles, and a depth camera captures the overall movement contour of the upper limb. This data is transmitted in real time to an intelligent assessment and feedback device and a virtual scenario construction and interaction device. The intelligent assessment and feedback device analyzes movement deviations, and the virtual scenario construction and interaction device adjusts the feedback from the virtual model in the scenario based on the patient's movements, allowing the patient to intuitively see the difference between their movements and the standard movements.

[0036] This process enables comprehensive and accurate collection of patient movement data. The combination of movement trajectory data and electromyographic signal data provides a rich data source for subsequent accurate assessment. Real-time acquisition and transmission ensure immediate interaction between the virtual scene and the patient's movements, enhancing the patient's immersion and experience. Simultaneously, automated data acquisition reduces human intervention and avoids data recording errors, providing reliable assurance for subsequent intelligent analysis and feedback, making rehabilitation training guidance and risk scenario experiences more targeted.

[0037] S30: The virtual scene construction and interaction device drives the branch evolution of the virtual scene based on the received motion data. The intelligent assessment and feedback device calculates the motion deviation degree through the motion deviation analysis model. The real-time feedback unit provides voice prompts and visual prompts to the patient through the augmented reality wearable terminal.

[0038] Specifically, after the virtual scene construction and interaction device receives motion data transmitted by the motion capture and data acquisition device, the interaction logic engine immediately analyzes the data. The interaction logic engine pre-defines branching evolution rules for the virtual scene, which correspond to the behavioral requirements within the scene. By comparing the received patient motion data with the pre-defined rules, the interaction logic engine determines whether the patient's action is correct or incorrect, thereby driving the virtual scene to evolve towards the corresponding branch. If the patient's action conforms to the scene's pre-define correct behavior rules, the virtual scene will display a screen showing successful risk avoidance or a training motion standard; if the patient's action conforms to the incorrect behavior rules, the virtual scene will display a screen showing the corresponding risk consequences or a prompt indicating non-standard action.

[0039] Simultaneously, the intelligent assessment and feedback device receives motion trajectory data and electromyographic signal data transmitted from the motion capture and data acquisition device. The motion deviation analysis model calls upon the standard motion parameter library stored in the cloud data management device to compare the patient's motion data with the standard motion parameters. Through the motion deviation calculation formula, the motion deviation analysis model accurately calculates the deviation of the patient's current motion. During the calculation process, the model automatically adjusts the motion trajectory weight coefficient and electromyographic signal weight coefficient according to the type of rehabilitation training to ensure that the deviation calculation result is highly consistent with the training goal.

[0040] Once the motion deviation analysis model determines the degree of motion deviation, the real-time feedback unit generates corresponding prompts based on the results. If the deviation does not exceed a preset threshold, the real-time feedback unit generates positive encouragement voice and visual prompts; if the deviation exceeds the preset threshold, the real-time feedback unit generates voice and visual prompts for motion correction, clearly informing the patient of the specific location and direction of adjustment for the motion deviation. These prompts are presented synchronously through the augmented reality wearable terminal. Voice prompts are played through the terminal's audio output unit, and visual prompts are displayed on a high-definition screen, ensuring that the patient can receive and respond to the prompts promptly.

[0041] Taking the education on the risk of falling from bed and upper limb muscle strength rehabilitation training for elderly hypertensive patients as an example, after entering the simulated scenario of falling from bed, the patient makes the action of getting out of bed without pulling up the bed rails. The motion capture and data acquisition device collects the motion data and transmits it to the virtual scene construction and interaction device. The interaction logic engine judges the action as an incorrect behavior and drives the scene to display the image of the patient's limb abrasions after falling from bed. At the same time, the intelligent assessment and feedback device calculates the deviation of the action, and the real-time feedback unit plays voice prompts through the augmented reality wearable terminal, informing the patient of the risk of not pulling up the bed rails, and at the same time displays the animation of the correct steps to get out of bed on the screen. In the subsequent upper limb muscle strength rehabilitation training, the electromyographic signal intensity of the patient's arm raising action is insufficient, and the motion deviation analysis model calculates that the deviation exceeds the threshold. The real-time feedback unit immediately prompts the patient to increase the force exerted and demonstrates the correct force exertion method through the virtual model.

[0042] This process enables real-time interaction between the virtual scene and the patient's actions, allowing the patient to intuitively experience the consequences of their behavior. Precise calculation of movement deviation provides a scientific basis for real-time feedback, ensuring the relevance and accuracy of the prompts. The real-time feedback mechanism helps patients correct incorrect movements promptly, deepening their memory of correct behaviors and standard actions. Through the synergistic work of virtual scene interaction and intelligent assessment feedback, the immersion and effectiveness of the educational process are significantly enhanced, allowing patients to proactively master risk avoidance methods and rehabilitation training techniques during the experience.

[0043] S40: After the education campaign ends, the education effectiveness evaluation model generates an education effectiveness evaluation report and uploads it to the cloud data management device; medical staff can view the evaluation report through medical staff terminals and push personalized supplementary education content to augmented reality wearable terminals.

[0044] Specifically, after the education session concludes, the education effectiveness evaluation model automatically starts working. This model retrieves various data stored throughout the entire education process, including the patient's correct choice rate in the risk simulation scenario, the average deviation of movements during rehabilitation training, and the score from the Q&A session that pops up after the education session. The model comprehensively analyzes this data according to preset evaluation rules, generating a multi-dimensional evaluation result of the education effectiveness. The evaluation result clearly classifies the patient's level of mastery of the education content and identifies the patient's weaknesses, such as insufficient risk awareness or non-standard execution of certain rehabilitation movements.

[0045] After the assessment is completed, the education effectiveness evaluation model automatically generates a standardized education effectiveness evaluation report. The evaluation report includes basic patient information, the name of the education scenario, data for various assessment indicators, mastery level, and analysis of weaknesses. Subsequently, the evaluation report is uploaded to a cloud data management device via wireless communication protocol. The cloud data management device categorizes and stores the report, links it with the patient's previous education records, and forms a complete personal education file, providing data support for subsequent education optimization.

[0046] Healthcare professionals can access patient education effectiveness evaluation reports stored in the cloud-based data management device at any time by logging into the healthcare terminal. The terminal's display interface presents the report content in a clear and intuitive format, allowing healthcare professionals to quickly grasp the patient's education progress. For weaknesses identified in the evaluation report, healthcare professionals, based on the patient's clinical condition and recovery progress, select corresponding supplementary education content from the scenario library and education resource library of the cloud-based data management device. Supplementary education content includes various forms such as targeted repetitive demonstrations of risk scenarios, breakdown teaching of rehabilitation movements, and reinforced explanations of key knowledge points.

[0047] Once the supplementary educational content is determined, healthcare staff issue a push command via their medical terminal. This command is transmitted to a cloud data management device via a wireless communication protocol. Upon receiving the command, the cloud data management device retrieves the corresponding supplementary educational content and pushes it to the patient's augmented reality wearable terminal. The next time the patient wears the augmented reality wearable terminal, the terminal will automatically prompt that new supplementary educational content is available, and the patient can choose to learn it immediately or later, depending on their own situation.

[0048] Taking a diabetic patient with lower extremity neuropathy requiring foot care education and lower extremity rehabilitation training as an example, after the education was completed, the evaluation report generated by the education effectiveness assessment model showed that the patient's correct selection rate for foot care-related risk scenarios was 65%, the average deviation of lower extremity rehabilitation training movements was 0.4, the question-and-answer score was 70 points, the mastery level was qualified, and the weaknesses were insufficient foot pressure perception and non-standard ankle flexion and extension movements. After reviewing the report through a medical terminal, medical staff selected a foot pressure perception enhancement risk simulation scenario and an ankle flexion and extension movement decomposition rehabilitation training scenario from the scenario library and pushed them to the patient's augmented reality wearable terminal. After wearing the terminal, the patient first deepened their understanding of foot care risks through the enhanced risk scenario, and then corrected their rehabilitation movements through the decomposition training scenario, achieving targeted improvement.

[0049] This process enables quantitative evaluation of educational effectiveness, allowing medical staff to clearly understand patients' actual understanding and avoiding the difficulty in measuring the effectiveness of traditional educational methods. The precise delivery of personalized supplementary educational content ensures the rational use of educational resources, allowing patients to focus on strengthening their weaknesses. Simultaneously, complete educational records provide a reliable basis for medical staff to adjust subsequent educational plans, forming a closed loop of educational evaluation, feedback, and optimization, continuously improving the accuracy and effectiveness of educational work.

[0050] In one embodiment, in step S30, the intelligent evaluation and feedback device achieves data fusion through a spatiotemporal synchronization calibration algorithm, including: using the action trigger time of the inertial measurement unit as the reference timestamp T0; detecting the signal trigger time T1 of the electromyography (EMG) sensor and calculating the time difference ΔT = T1 - T0; shifting the EMG signal along the time axis according to the difference between ΔT and ΔT0 to achieve time alignment between the EMG signal and the action trajectory data; and performing spatiotemporal correlation fusion of the aligned EMG signal and the action trajectory data; where ΔT0 is the EMG-action synchronization time difference of the standard action. Specifically, the intelligent evaluation and feedback device achieves data fusion through a spatiotemporal synchronization calibration algorithm, with the core objective of eliminating the time delay between the action trajectory data collected by the inertial measurement unit and the EMG signal data collected by the EMG sensor, achieving precise spatiotemporal alignment of the two types of data without changing the intensity or amplitude attributes of any original signal.

[0051] When the algorithm is executed, the trigger time of the inertial measurement unit's action is first used as the reference timestamp T0. The inertial measurement unit detects the moment when the patient begins to perform the preset action and immediately records that moment as T0, providing a unified reference benchmark for the time calibration of all data.

[0052] The signal trigger time T1 of the electromyography (EMG) sensor is detected synchronously, which is the moment when the EMG sensor captures the first effective electrical signal generated by the target muscle group. The time difference ΔT is calculated, which reflects the actual time delay between the triggering of the action and the start of the EMG signal.

[0053] The depth camera continuously captures a sequence of patient movement frames. Each frame contains a corresponding time stamp and movement posture information, accurately reconstructing the complete time process of the movement. The electromyographic synchronization time difference ΔT0 of the standard movement stored in the cloud data management device is an ideal time delay range derived from extensive clinical trials, representing the normal range of natural synchronization between movement and electromyographic signals.

[0054] The calibration process aligns only the time dimension and does not involve scaling the electromyographic signal intensity. When ΔT is within the allowable error range of ΔT0, the time stamps of the original movement trajectory data and electromyographic signal data are directly retained; when ΔT exceeds the error range of ΔT0, the time stamps of the electromyographic signal are corrected by time axis translation. Specifically, the entire segment of electromyographic signal acquired by the electromyographic sensor is time-shifted according to the difference between ΔT and ΔT0, so that key nodes such as the start time and peak time of the electromyographic signal completely coincide with the corresponding key nodes of the movement trajectory data in time.

[0055] In the spatial dimension, by combining the motion posture features captured by the depth camera, the aligned electromyographic signals and motion trajectory data are associated one by one according to the timestamp, ensuring that the electromyographic signal data corresponding to each motion posture is accurately matched, and realizing the spatiotemporal synchronous fusion of the two types of data.

[0056] Taking elbow flexion and extension rehabilitation training as an example, the electromyographic (EMG) synchronization time difference ΔT0 for standard movements is 3 milliseconds, meaning that under normal circumstances, the EMG signal should start 3 milliseconds after the movement is triggered. In actual data acquisition, the T0 recorded by the inertial measurement unit and the T1 recorded by the EMG sensor resulted in a ΔT of 5 milliseconds, exceeding the error range. The system shifted the entire EMG signal forward by 2 milliseconds to align the start time of the EMG signal with the 3-millisecond time after the movement is triggered. Simultaneously, it correlated the elbow flexion and extension posture data captured by the depth camera to complete the spatiotemporal synchronization calibration.

[0057] In one embodiment, such as Figure 2 As shown, a patient education system based on augmented reality scene simulation is provided. This patient education system based on augmented reality scene simulation corresponds one-to-one with the patient education method based on augmented reality scene simulation in the above embodiments. The patient education system based on augmented reality scene simulation includes: an augmented reality wearable terminal, a motion capture and data acquisition device, a virtual scene construction and interaction device, an intelligent assessment and feedback device, a cloud data management device, and a medical terminal. Each device exchanges data through a wireless communication protocol. The functional modules are described in detail below: Augmented reality wearable devices are used to present virtual scenes and receive voice commands from patients; The motion capture and data acquisition device is used to collect patient movement trajectory data and electromyographic signal data and transmit them to the intelligent assessment and feedback device and the virtual scene construction and interaction device. The virtual scene construction and interaction device is used to construct risk simulation sub-scenes and rehabilitation training sub-scenes, and drives the evolution of scene branches based on the received patient action data; The intelligent assessment and feedback device is used to analyze movement deviations and generate assessment results, providing feedback to patients through an augmented reality wearable terminal; The cloud-based data management device is used to store data and update the scene and parameter libraries; the medical terminal is used to view the evaluation results and push personalized educational content to the augmented reality wearable terminal.

[0058] Specifically, the augmented reality wearable terminal is the core carrier for patients to interact with the virtual scene. Its high-definition display screen can clearly present the dynamic images of the risk simulation sub-scene and rehabilitation training sub-scene, allowing patients to obtain an immersive visual experience. The terminal's voice interaction unit continuously listens to the patient's voice commands. The patient can directly issue commands such as repeating the demonstration, adjusting the scene speed, and pausing the experience. The terminal responds quickly after receiving the commands and flexibly adjusts the scene's operating status, improving the ease of operation for the patient. The positioning unit captures the patient's position changes in physical space in real time, ensuring that the virtual scene and the real environment always remain spatially synchronized, avoiding scene misalignment caused by positional shifts, and ensuring the continuity of interaction.

[0059] The motion capture and data acquisition device is responsible for acquiring real-time patient behavior data. The inertial measurement unit accurately collects angular velocity and acceleration data of the patient's head posture and limb movements, completely recording the trajectory changes and amplitude of the movements. Electromyography (EMG) sensors closely adhere to the target muscle groups corresponding to the rehabilitation training, continuously capturing the intensity of electrical signals generated by muscle contractions, reflecting the force exerted by the patient's movements and the state of muscle activation. A depth camera captures the patient's full-body movement contours from a global perspective, transforming two-dimensional movements into a three-dimensional movement model, achieving a comprehensive record of the patient's overall movements. This acquired motion trajectory data and EMG signal data are synchronously transmitted via wireless communication protocols to the intelligent assessment and feedback device and the virtual scene construction and interaction device, providing raw data support for subsequent analysis and interaction.

[0060] The virtual scene construction and interaction device has a built-in scene library and interaction logic engine. The scene library pre-stores various risk simulation sub-scenes and rehabilitation training sub-scenes. The risk simulation sub-scenes cover common adverse event scenarios such as falls, bed falls, and dislodgement, with each scene having two preset interaction branches: correct behavior and incorrect behavior. The rehabilitation training sub-scenes construct standardized virtual models of rehabilitation movements for different diseases and rehabilitation stages, and set corresponding standard movement parameter thresholds. The interaction logic engine receives patient movement data transmitted from the motion capture and data acquisition device, compares it with the preset behavioral logic and parameter thresholds of the scene, determines the attributes of the patient's movements, and then drives the virtual scene to evolve towards the corresponding branch, allowing the patient's actions to directly affect the scene's outcome.

[0061] The intelligent assessment and feedback device is the core of achieving accurate assessment and real-time guidance. The movement deviation analysis model receives movement trajectory data, electromyographic signal data, and standard movement parameter library data. Through a movement deviation calculation formula combined with dynamic weight adjustment, it accurately calculates the degree of deviation in the patient's movements. The education effectiveness evaluation model integrates multi-dimensional indicators such as scene interaction accuracy rate, mean movement deviation rate, and education question-and-answer accuracy rate to generate the patient's mastery level of the education content and identify weaknesses. Based on the movement deviation analysis results and assessment conclusions, the real-time feedback unit generates voice and visual prompts, which are delivered to the patient through an augmented reality wearable terminal to promptly correct incorrect movements and reinforce correct cognition.

[0062] As the core platform for data storage and resource updates, the cloud-based data management device is responsible for storing various data, including basic patient information, action data collection, and evaluation results of patient education effectiveness, thus constructing a complete patient education file. Simultaneously, the cloud-based data management device continuously maintains and updates the risk scenario library and rehabilitation training action parameter library. Based on new data and needs accumulated in clinical practice, it supplements new scenario resources and optimizes the parameter settings of existing scenarios, ensuring the timeliness and applicability of the scenario and parameter libraries. Through big data analytics, the cloud-based data management device can also identify weaknesses in patient education for different patient groups, providing data support for optimizing overall patient education content.

[0063] The healthcare terminal serves as an operational platform for healthcare professionals to conduct patient education and management. Through the terminal, healthcare professionals can access patient education effectiveness evaluation reports stored in the cloud-based data management device at any time, clearly understanding the progress and weaknesses of education for each patient. Based on these evaluation reports, and considering the patient's clinical condition and rehabilitation plan, healthcare professionals select personalized supplementary education content from the scenario library and education resource library of the cloud-based data management device. They then issue push commands through the terminal to deliver this content to the patient's augmented reality wearable device, achieving targeted and enhanced education.

[0064] Taking rehabilitation education for elderly stroke patients as an example, after the patient wears an augmented reality wearable terminal, the terminal presents a simulated fall risk scenario in the ward and an upper limb rehabilitation training scenario. When the patient attempts to get up in the scenario, the motion capture and data acquisition device collects the trajectory data of the getting-up movement and the electromyographic signal data of the biceps brachii muscle, and transmits it to the intelligent assessment and feedback device and the virtual scene construction and interaction device. The virtual scene construction and interaction device determines that the patient is getting up too quickly and drives the scenario to display the consequences of dizziness and falling. The intelligent assessment and feedback device calculates the movement deviation and prompts the patient to get up slowly through the terminal. After the education is completed, the assessment report is uploaded to the cloud data management device. Medical staff view the report through the medical staff terminal and find that the patient's awareness of fall risk is insufficient, and push fall prevention reinforcement scenarios to the terminal.

[0065] This system architecture enables collaborative operation among various devices. Augmented reality wearable terminals provide an immersive experience and convenient interaction; motion capture and data acquisition devices ensure comprehensive and accurate data collection; virtual scene construction and interaction devices enable dynamic scene responses; intelligent assessment and feedback devices provide personalized guidance and quantitative evaluation; cloud data management devices ensure data storage and resource updates; and medical terminals support precise management by medical staff. Each device performs its specific function while working closely together to form a complete educational loop, effectively enhancing the immersiveness, relevance, and effectiveness of educational outreach, while reducing the workload of medical staff and providing scientific support for patients' risk avoidance and rehabilitation training.

[0066] Optionally, the augmented reality wearable terminal includes a high-definition display screen, a voice interaction unit, a positioning unit, a haptic feedback unit, and an environmental sound effect simulation unit; the high-definition display screen is used to present virtual scenes and virtual models for rehabilitation training; the voice interaction unit is used to receive voice commands from the patient; the positioning unit is used to obtain the patient's physical spatial location information; the haptic feedback unit is used to generate vibration feedback at the corresponding limbs of the patient; and the environmental sound effect simulation unit is used to play environmental sound effects corresponding to the virtual scene.

[0067] Specifically, the augmented reality wearable terminal integrates a high-definition display screen, a voice interaction unit, a positioning unit, a haptic feedback unit, and an environmental sound simulation unit. These units work together to provide patients with a multi-sensory immersive educational experience. All units are integrated into a lightweight terminal body, with a reasonable layout and comfortable wear, without hindering the patient's movements, making it suitable for patients of different ages and physical conditions.

[0068] The high-definition display screen utilizes high-resolution flexible display technology and is installed to fit the patient's visual angle. Its image is clear and detailed, accurately presenting the dynamic details of risk simulation sub-scenes and the standard movements of the virtual rehabilitation training model. Whether it's the changes in ground texture in a fall scenario or the demonstration of muscle contraction and relaxation in rehabilitation training, everything can be presented intuitively on the screen. The screen's brightness and contrast automatically adjust according to ambient light, ensuring that patients can clearly see the scene content under different lighting conditions, avoiding the impact of strong or weak light environments on the visual experience.

[0069] The voice interaction unit features a built-in high-sensitivity microphone and voice recognition module, enabling accurate capture of patients' voice commands. Patients do not need to manually operate the system; they can issue control commands simply by speaking naturally. Common commands include repeating the current scene, pausing the educational session, adjusting the playback speed, and switching scenes. The voice recognition module is interference-resistant, filtering out environmental noise, accurately recognizing patient commands, and responding quickly, allowing patients to flexibly control the educational process and improving ease of use.

[0070] The positioning unit employs spatial positioning technology to capture the patient's position coordinates and posture changes in physical space in real time. The positioning process is continuous and accurate, capable of tracking the patient's movement trajectory and body rotation in real time. As the patient moves in physical space, the positioning unit transmits the location information to the virtual scene construction and interaction device in real time. The device adjusts the presentation perspective and range of the virtual scene according to the position change, ensuring that the virtual scene remains spatially synchronized with the real environment. Even with slight movements or turns by the patient, the virtual scene adapts seamlessly, avoiding scene misalignment or disconnection and ensuring the continuity of immersion.

[0071] The haptic feedback unit consists of multiple miniature vibration modules distributed across key areas of the patient's limbs. When a patient makes a mistake or faces risk in a virtual scenario, the haptic feedback unit vibrates at the corresponding limb area. The intensity and frequency of the vibration are dynamically adjusted based on the scenario content, matching the intensity of events in the virtual environment. For example, in a fall risk scenario, when a patient simulates falling from a bed, the vibration module corresponding to the point of impact will generate continuous and strong vibrations; in a rehabilitation training scenario, when insufficient force is exerted, the vibration module corresponding to the muscle group will generate slight and continuous vibrations, reminding the patient to increase effort.

[0072] The environmental sound simulation unit has a built-in high-quality audio output module that can play environmental sound effects corresponding to the virtual scene. Different scenes are equipped with exclusive sound effects. In the risk simulation scene, environmental sounds related to the event will be played, while in the rehabilitation training scene, clear guiding sounds and confirmation sounds after the action is completed will be played. The volume of the sound effects can be adjusted according to the patient's needs. The audio output is clear and free of noise, enhancing the realism of the scene through auditory stimulation, making it easier for patients to immerse themselves in the scene atmosphere and deepen their memory of the educational content.

[0073] Taking post-operative rehabilitation patient education as an example, patients wear augmented reality wearable terminals for lower limb rehabilitation training and fall risk avoidance education. A high-definition display clearly shows a virtual model of knee flexion and extension rehabilitation training and fall risk scenarios in the ward, allowing patients to clearly see the details of standard movements and the process of a fall. When a patient wants to repeat the demonstration, they issue a voice command to repeat it, and the voice interaction unit quickly recognizes and controls the scene to replay. As the patient moves within the scene, the positioning unit tracks their position in real time, and the virtual scene adjusts its perspective accordingly, always maintaining synchronization with the real environment. When a patient makes the mistake of not holding onto a handrail in a fall scenario, the haptic feedback unit vibrates on the patient's arm, while the environmental sound simulation unit plays a collision sound, allowing the patient to intuitively experience the risk.

[0074] The various units of the augmented reality wearable terminal work collaboratively to construct an immersive experience environment from multiple dimensions, including vision, hearing, and touch. A high-definition display ensures clear presentation of scene content, a voice interaction unit enhances ease of operation, a positioning unit ensures synchronization between the scene and the environment, and a haptic feedback unit and an environmental sound simulation unit enhance the realism and impact of the scene. This multi-sensory experience makes it easier for patients to engage in the educational process, deeply perceive the risks and consequences and standard operating procedures, significantly improving the effectiveness of the education. At the same time, the terminal's lightweight design and convenient operation lower the barrier to entry for patients, adapting to the needs of different patients and supporting the widespread implementation of educational work.

[0075] Optionally, the motion capture and data acquisition device includes an inertial measurement unit, an electromyography (EMG) sensor, and a depth camera; the inertial measurement unit is used to acquire angular velocity and acceleration data of the patient's head posture and limb movements; the EMG sensor is used to acquire the EMG signal intensity of the target muscle groups in rehabilitation training; and the depth camera is used to capture the contours of the patient's whole-body movements and generate a three-dimensional motion model.

[0076] Specifically, the motion capture and data acquisition device consists of an inertial measurement unit, an electromyography (EMG) sensor, and a depth camera. These three components work together to achieve comprehensive and accurate acquisition of patient movements and physiological signals. The device's components are rationally arranged: the inertial measurement unit is integrated into the augmented reality wearable terminal, the EMG sensor is an adhesive design, and the depth camera is deployed in a fixed position in the education area, ensuring that it does not interfere with the patient's movement while guaranteeing full coverage of the acquisition range.

[0077] The inertial measurement unit (IMU) incorporates angular velocity and accelerometer sensors to monitor the patient's head posture and limb movements in real time. When the patient performs actions such as turning their head, turning around, and extending or bending their limbs in a virtual environment, the angular velocity sensor accurately records the rotational speed, while the accelerometer captures the rate of change of velocity. This data can completely reconstruct the trajectory, amplitude, and rhythm of the patient's movements, providing fundamental data for subsequent analysis of whether the movements are performed correctly. The IMU's high data acquisition frequency allows it to capture subtle changes in movement, ensuring data continuity and integrity.

[0078] The electromyography (EMG) sensor is a flexible, adhesive structure that fits snugly onto the target muscle group corresponding to the rehabilitation training. During rehabilitation exercises, muscle contractions generate weak electrical signals, which the EMG sensor precisely captures. Different rehabilitation movements target different muscle groups, and medical staff attach the EMG sensor to the appropriate locations based on the training content. For example, it is attached to the biceps and triceps brachii for upper limb rehabilitation and to the quadriceps and gastrocnemius for lower limb rehabilitation. The EMG sensor continuously collects electrical signal data, which directly reflects the force exerted by the patient's movements and the degree of muscle activation, serving as a crucial basis for evaluating training effectiveness.

[0079] Depth cameras are deployed above or to the side of the educational area, covering the entire area where the patient is moving. These cameras acquire depth information of the environment by emitting and receiving infrared light, continuously capturing the patient's full-body movement contours as they perform actions. Each frame contains the positional relationships and morphological features of the patient's limbs. The system processes this image data, extracts key motion nodes, and generates a 3D motion model synchronized with the patient's movements. This 3D motion model visually displays the patient's posture, facilitating comparison with a standard motion model and quickly identifying the location of movement deviations.

[0080] Taking upper limb lifting rehabilitation training for stroke patients as an example, an inertial measurement unit (IMU) is integrated into the arm of an augmented reality wearable terminal to collect real-time angular velocity and acceleration data of the patient's lifting movements, recording changes in speed and amplitude. An electromyography (EMG) sensor is attached to the surface of the patient's biceps to capture the intensity of electrical signals during muscle contraction, reflecting the force exertion. A depth camera is deployed to the side of the rehabilitation training area to continuously capture the patient's full-body movement contours, generating a three-dimensional movement model of the upper limb lifting. The data collected by these three devices is synchronously transmitted to an intelligent assessment and feedback device, providing comprehensive data support for movement deviation analysis.

[0081] The motion capture and data acquisition device comprehensively records patient movements and physiological signals through multi-dimensional data acquisition. An inertial measurement unit captures dynamic changes in movement, an electromyography (EMG) sensor reflects muscle exertion, and a depth camera presents the overall posture. These three types of data complement and verify each other, ensuring the comprehensiveness and accuracy of the collected data. Precise data acquisition provides a reliable foundation for subsequent movement deviation analysis and educational effectiveness evaluation, making the assessment results more scientific and convincing. At the same time, the device's design does not interfere with the patient, ensuring a comfortable experience during the educational process and improving patient cooperation and the smooth implementation of educational work.

[0082] Optionally, the virtual scene construction and interaction device includes a scene library and an interaction logic engine; the scene library contains risk simulation sub-scenes and rehabilitation training sub-scenes, the risk simulation sub-scenes preset correct behavior interaction branches and incorrect behavior interaction branches, and the rehabilitation training sub-scenes preset standard action parameter thresholds; the interaction logic engine is used to receive action data transmitted by the motion capture and data acquisition device, determine whether the patient's actions conform to the preset logic of the scene, and drive the evolution of virtual scene branches.

[0083] Specifically, the core components of the virtual scene construction and interactive device are the scene library and the interaction logic engine. These two work together to achieve accurate presentation and dynamic interaction of the virtual scene, providing patients with an immersive educational experience tailored to clinical needs. The scene library is the core storage for pre-set content, while the interaction logic engine is the core control for scene responses. The two work closely together through data transmission to ensure that the scene can be adjusted in real time according to the patient's actions.

[0084] The scenario library categorizes and stores risk simulation sub-scenarios and rehabilitation training sub-scenarios. All scenarios are built based on common clinical situations and rehabilitation medicine standards, ensuring the professionalism and practicality of the content. Risk simulation sub-scenarios cover common adverse events during hospitalization, such as falls, bed falls, and dislodgement. Each risk simulation sub-scenario has two parallel interaction branches: a correct behavior interaction branch and an incorrect behavior interaction branch. The correct behavior interaction branch corresponds to standard operating procedures that can avoid risks, while the incorrect behavior interaction branch corresponds to common risky behaviors in clinical practice and their subsequent consequences. Rehabilitation training sub-scenarios are designed for different diseases and different rehabilitation stages, covering various types such as joint mobility, muscle strength training, and balance training. Each rehabilitation training sub-scenario has preset standard movement parameter thresholds, including the angle range, speed range, force intensity standards, etc., providing a clear basis for movement assessment.

[0085] The interaction logic engine is crucial for achieving real-time interaction between the scene and the patient's actions. It continuously receives patient movement data from motion capture and data acquisition devices, including movement trajectory data, electromyographic signal data, and 3D movement model data. Upon receiving the data, the engine compares the patient's movement data with the preset logic of the current scene. For risk simulation sub-scenes, the engine determines whether the patient's action is correct or incorrect, triggering the corresponding scene branch. For rehabilitation training sub-scenes, the engine compares the patient's movement data with preset standard movement parameter thresholds to determine if the action meets the standard requirements.

[0086] After the comparison is complete, the interaction logic engine immediately drives the evolution of the virtual scene branches. If the patient performs the correct behavior in the risk simulation sub-scene, the scene will display the result of successful risk avoidance and present key points of the correct behavior; if the patient performs the wrong behavior, the scene will dynamically demonstrate the occurrence of the risk event and its subsequent adverse consequences, while providing an analysis of the cause of the error. In the rehabilitation training sub-scene, if the patient's movements meet the standard movement parameter thresholds, the scene will maintain the current training pace and provide positive guidance; if the movements exceed the threshold range, the scene will pause or slow down, highlighting the deviation areas to provide scene support for subsequent real-time feedback.

[0087] Taking the education on the risk of falling out of bed for elderly patients as an example, the risk simulation sub-scenario has two preset branches. The correct behavior interaction branch requires the patient to sit up for three seconds before slowly getting out of bed and holding onto the handrail. The incorrect behavior interaction branch requires the patient to get out of bed quickly without holding onto the handrail. After the patient enters the scene wearing an augmented reality wearable terminal, the motion capture and data acquisition device collects data such as the patient's speed of getting up and limb movements and transmits it to the interaction logic engine. If the patient gets out of bed quickly, the interaction logic engine determines that this action is incorrect and drives the scene to display dynamic images of the patient's limb abrasions after falling out of bed and the scene of medical staff handling the situation. If the patient gets out of bed according to the correct procedure, the interaction logic engine triggers the correct behavior branch, displays the result of getting out of bed safely, and pops up a reminder of the importance of holding onto the handrail when getting up.

[0088] In the knee flexion and extension rehabilitation training sub-scenario, the preset standard motion parameter thresholds are a flexion and extension angle of 0 to 150 degrees and a motion speed of 30 degrees per second. During training, the interaction logic engine receives angle and speed data transmitted from the motion capture and data acquisition device and compares it with the preset thresholds. If the patient's knee flexion and extension angle reaches 160 degrees, the interaction logic engine determines that the motion exceeds the threshold, pauses the scene, and highlights the knee joint area, providing deviation position information to the real-time feedback unit.

[0089] This design enables dynamic responses in the virtual scenario, allowing patient actions to directly determine the scenario's progression, breaking the limitations of traditional fixed-playback-style educational methods. The dual-branch design of the risk simulation sub-scenario allows patients to intuitively experience the differences in consequences of different behaviors, deepening their understanding of risk. The standard parameter threshold settings for the rehabilitation training sub-scenario provide a clear basis for action assessment, ensuring consistency between scenario evolution and action evaluation. Through the collaborative work of the scenario library and the interactive logic engine, the virtual scenario can accurately match patient behavior, significantly enhancing the immersion and relevance of the educational process, helping patients quickly master risk avoidance methods and standard rehabilitation movements.

[0090] Optionally, the intelligent assessment and feedback device includes a movement deviation analysis model, a health education effectiveness evaluation model, and a real-time feedback unit. The movement deviation analysis model receives three-dimensional movement data, electromyographic signal data, and standard movement parameter library data from the patient, and calculates the movement deviation degree using a movement deviation degree calculation formula. The health education effectiveness evaluation model combines the accuracy rate of scene interaction, the average value of movement deviation degree, and the accuracy rate of health education questions and answers to generate the patient's mastery level. The real-time feedback unit provides movement correction information to the patient through voice prompts and visual prompts.

[0091] Specifically, the intelligent assessment and feedback device consists of a movement deviation analysis model, an education effectiveness evaluation model, and a real-time feedback unit. These three components work together to accurately analyze movement data, comprehensively evaluate the education effectiveness, and provide timely, targeted feedback, offering personalized education guidance to patients. The device receives multi-source data and performs hierarchical processing to ensure the scientific validity of the assessment results and the accuracy of the feedback content.

[0092] The core function of the movement deviation analysis model is to quantify the difference between a patient's movements and standard movements. The model receives real-time 3D movement data and electromyographic (EMG) signal data from motion capture and data acquisition devices, while simultaneously accessing a standard movement parameter library stored in a cloud-based data management system. The 3D patient movement data includes spatial features such as the angle, trajectory, and speed of limb movements; the EMG signal data reflects the force intensity of the target muscle group; and the standard movement parameter library provides ideal reference indicators for corresponding training movements. The model compares the patient data with the standard data dimension by dimension, quantifying the degree of deviation using a movement deviation calculation formula. During the calculation, the model dynamically adjusts the weighting coefficients of the movement trajectory and EMG signals based on the type of rehabilitation training, ensuring the relevance of the deviation calculation under different training objectives and making the assessment results more aligned with clinical training needs.

[0093] The education effectiveness evaluation model focuses on comprehensively assessing patients' mastery of the educational content. The model collects three core data points: the patient's accuracy rate in scenario interactions within a risk simulation sub-scenario, the average deviation rate of movements during rehabilitation training, and the accuracy rate of questions and answers during the post-education Q&A session. Scenario interaction accuracy reflects the patient's ability to identify and avoid risky behaviors; the average deviation rate of movements reflects the standardization of rehabilitation training; and the accuracy rate of questions and answers directly measures the patient's memory and understanding of core knowledge points. The model weights and calculates the three types of data according to preset evaluation rules, classifying patients' mastery levels into different categories based on the calculation results. It also precisely identifies patients' weaknesses, such as insufficient risk awareness, non-standard execution of specific movements, or weak memory of knowledge points.

[0094] The real-time feedback unit is responsible for transforming the analysis and evaluation results into patient-perceptible prompts. The unit receives movement deviation data from the movement deviation analysis model and preliminary judgments from the education effectiveness evaluation model, generating corresponding voice and visual prompts. When the movement deviation does not exceed a preset threshold, the real-time feedback unit generates positive reinforcement prompts to encourage correct behavior; when the deviation exceeds the threshold, the unit generates movement correction prompts, clearly informing the patient of the specific location of the deviation, the direction of adjustment, and the required standards. Voice prompts are clearly played through the audio output unit of the augmented reality wearable terminal, while visual prompts are presented on a high-definition display screen in the form of text, animation, or highlighted markers, ensuring that patients can quickly understand and adjust their movements promptly.

[0095] Taking a patient undergoing lumbar spine surgery requiring back muscle rehabilitation training as an example, when the patient performs the "Little Swallow" rehabilitation exercise, the movement deviation analysis model receives the extension angle and movement speed of the back muscles from the three-dimensional data of the movement, and the electrical signal intensity of the back muscles from the electromyographic data. It then calls upon the angle range, speed standard, and force intensity reference values ​​for that movement from the standard movement parameter library. The movement deviation is calculated using a formula. If the patient's back muscle extension angle is insufficient, the model determines that the deviation exceeds the standard. After the patient completes multiple training sessions, the education effectiveness evaluation model combines the patient's interaction accuracy in fall risk scenarios, the average deviation of multiple movements, and the correctness of answering questions about back muscle protection knowledge points to generate an assessment result indicating a passing level of mastery. Simultaneously, it marks the weak point as insufficient back muscle extension angle. The real-time feedback unit then plays voice prompts through the augmented reality wearable terminal, informing the patient that the extension angle is insufficient, and simultaneously demonstrates the extension range of the standard movement in animation form on the display screen, guiding the patient to adjust.

[0096] This design achieves quantification of movement assessment and immediacy of feedback. The movement deviation analysis model, through multi-source data fusion and dynamic weight adjustment, makes deviation assessment more accurate and reliable. The multi-dimensional indicator integration of the education effectiveness assessment model comprehensively reflects the patient's mastery of education, avoiding the limitations of a single assessment dimension. The real-time feedback unit ensures that patients receive corrective information promptly and improve their movements quickly through multi-sensory prompts. The collaborative work of these three components not only helps patients optimize their movements in real time during training but also provides medical staff with clear assessment criteria, significantly improving the standardization of rehabilitation training and the effectiveness of education.

[0097] Optionally, the formula for calculating the motion deviation is D=α×(1-cosθ)+β×(1-S p / S s ); where D is the movement deviation; α is the movement trajectory weighting coefficient; θ is the angle between the patient's movement trajectory vector and the standard movement trajectory vector; β is the electromyographic signal weighting coefficient; S p S represents the average actual electromyographic signal intensity of the patient's muscle groups. s This represents the average electromyographic signal intensity of the muscle group corresponding to the standard movement.

[0098] Specifically, the formula for calculating movement deviation is the core algorithm of the movement deviation analysis model, used to accurately quantify the degree of deviation between a patient's movements and standard movements. This formula integrates two key dimensions—movement trajectory deviation and electromyographic signal deviation—to achieve a comprehensive assessment of rehabilitation training movements. The calculation results directly provide data support for real-time feedback and the evaluation of educational effectiveness.

[0099] Each parameter in the formula has a clear physical meaning and data source. The movement deviation is a comprehensive assessment result, ranging from 0 to 1. A smaller value indicates a smaller deviation between the patient's movement and the standard movement, while a larger value indicates a larger deviation. The movement trajectory weighting coefficient is used to adjust the proportion of movement trajectory deviation in the total deviation, and the electromyography (EMG) signal weighting coefficient is used to adjust the proportion of EMG signal deviation in the total deviation. The sum of the two is 1, ensuring a reasonable weighting of the assessment dimensions.

[0100] The angle between the patient's motion trajectory vector and the standard motion trajectory vector is a core parameter reflecting motion trajectory deviation. The motion deviation analysis model processes the patient's three-dimensional motion data, extracts key nodes of the motion trajectory, and constructs the patient's motion trajectory vector. Simultaneously, it calls upon standard motion trajectory data from a standard motion parameter library to construct a standard motion trajectory vector. The angle between the two vectors is calculated using a vector angle calculation method. A smaller angle indicates that the patient's motion trajectory is closer to the standard trajectory, while a larger angle indicates a more significant trajectory deviation.

[0101] The average actual electromyographic (EMG) signal intensity of the patient's muscle groups is the average value of all valid EMG signals collected by the EMG sensor during a single complete rehabilitation movement. During data acquisition, the EMG sensor continuously captures the electrical signals of the target muscle groups. The system automatically filters out interference signals, retains valid data, and calculates the average value, which directly reflects the actual force exerted by the patient's movement. The average EMG signal intensity of the muscle groups corresponding to the standard movement is the average level of EMG signal intensity that the target muscle groups should achieve during the execution of the standard movement, stored in the cloud-based data management device. This value is derived from extensive clinical trials and statistical data, and is authoritative and valuable for reference.

[0102] The calculation process begins with calculating the movement trajectory deviation term. This is achieved by subtracting the cosine of the angle between the patient's movement trajectory vector and the standard movement trajectory vector from 1, resulting in the movement trajectory deviation coefficient. Multiplying this coefficient by the movement trajectory weighting coefficient yields the deviation contribution value for the movement trajectory dimension. Next, the electromyography (EMG) signal deviation term is calculated. This is achieved by subtracting the ratio of the average actual EMG signal intensity of the patient's muscle group to the average EMG signal intensity of the corresponding muscle group in the standard movement from 1, resulting in the EMG signal deviation coefficient. Multiplying this coefficient by the EMG signal weighting coefficient gives the deviation contribution value for the EMG signal dimension. Finally, the deviation contribution values ​​for both dimensions are summed to obtain the final movement deviation degree.

[0103] Taking elbow flexion and extension rehabilitation training as an example, the focus of this training is the accuracy of the movement trajectory. The weighting coefficient for the movement trajectory is 0.8, and the weighting coefficient for the electromyographic signal is 0.2. The angle between the standard movement trajectory vector and the patient's movement trajectory vector is 10 degrees. Subtracting the cosine of this angle from 1 yields 0.015. The average actual electromyographic signal intensity of the patient's muscle groups is 180 microvolts, while the average electromyographic signal intensity of the corresponding muscle groups in the standard movement is 200 microvolts. Subtracting the ratio of the two from 1 yields 0.1. According to the formula, the deviation contribution of the movement trajectory dimension is 0.8 multiplied by 0.015, which equals 0.012. The deviation contribution of the electromyographic signal dimension is 0.2 multiplied by 0.1, which equals 0.02. The final movement deviation is 0.012 plus 0.02, which equals 0.032. This value indicates that the patient's movement deviation is small and meets the training requirements.

[0104] If the patient exerts insufficient force during the training, the average electromyographic signal intensity is only 100 microvolts, while other parameters remain unchanged. The electromyographic signal deviation coefficient is 1 minus 100 divided by 200, which equals 0.5. The deviation contribution of the electromyographic signal dimension is 0.2 multiplied by 0.5, which equals 0.1. The movement deviation is 0.012 plus 0.1, which equals 0.112. At this point, the deviation increases, and the system will trigger real-time feedback.

[0105] This calculation formula enables the quantitative assessment of movement deviations. Through the weighted fusion of two core dimensions, it comprehensively considers the accuracy of movement trajectory and the standardization of muscle exertion. The weighting coefficients allow the assessment to be adapted to different types of rehabilitation training, meeting the assessment needs of different training goals. Precise quantitative results provide clear evidence for real-time feedback, enabling patients to clearly understand the degree of their movement deviations and providing objective assessment data for medical staff. Compared to traditional qualitative assessment methods, this formula makes movement assessment more scientific and objective, effectively improving the accuracy of rehabilitation training guidance and the reliability of educational outcomes.

[0106] Optionally, the motion deviation analysis model is configured with a dynamic weight coefficient adjustment function to adjust the values ​​of α and β based on the type of rehabilitation training; when the type of rehabilitation training is joint mobility training, α is 0.8 and β is 0.2; when the type of rehabilitation training is muscle strength training, α is 0.3 and β is 0.7; the weight coefficients can be manually adjusted through the medical terminal.

[0107] Specifically, the motion deviation analysis model has a built-in dynamic weight coefficient adjustment function. The core of this function is to automatically adapt the values ​​of the motion trajectory weight coefficient and the electromyographic signal weight coefficient according to different types of rehabilitation training. At the same time, it supports medical staff to manually intervene and adjust the values ​​through medical terminals to ensure that the motion deviation calculation is consistent with the core objectives of different training.

[0108] Rehabilitation training is broadly categorized into joint mobilization training and muscle strength training, with fundamentally different assessment focuses for each. The core objective of joint mobilization training is to restore the range of motion and the regularity of movement trajectories; therefore, the assessment should emphasize the accuracy of the movement trajectory. Based on this requirement, when the training type is joint mobilization training, the model automatically sets the weight coefficient for movement trajectory to 0.8 and the weight coefficient for electromyography (EMG) signals to 0.2, ensuring that movement trajectory deviation dominates the total deviation and that the assessment results focus on trajectory regularity.

[0109] The core objective of strength training is to enhance the contractile strength and force stability of the target muscle group, and the assessment should focus on whether the muscle force intensity meets the standard. Therefore, when the training type is strength training, the model adjusts the weight coefficients in reverse, setting the weight coefficient for the movement trajectory to 0.3 and the weight coefficient for the electromyographic signal to 0.7, so that the deviation of the electromyographic signal becomes the main component of the total deviation, allowing the assessment results to accurately reflect the force exertion.

[0110] In addition to automatic adjustments, the model allows healthcare professionals to manually adjust the weighting coefficients via a medical terminal. Healthcare professionals can flexibly modify the values ​​of the motion trajectory weighting coefficient and the electromyography signal weighting coefficient based on the patient's individual circumstances, such as limited joint movement but normal muscle strength, or extremely weak muscle strength but within acceptable range of motion. The adjustment is performed through a visual interface on the medical terminal, and the modified data is synchronized to the motion deviation analysis model in real time and immediately applied to subsequent deviation calculations.

[0111] Taking stroke rehabilitation patients as an example, patients initially need to perform shoulder joint mobility training to restore the range of motion. At this time, the model determines the training type as joint mobility training and automatically sets the motion trajectory weight coefficient to 0.8 and the electromyography signal weight coefficient to 0.2. During training, even slight deviations in the motion trajectory will be reflected in the total deviation, and the real-time feedback unit will provide targeted prompts to adjust the trajectory, helping patients to standardize joint movements.

[0112] When patients enter the later stage of muscle strength training, focusing on improving deltoid muscle strength, the model automatically adjusts the weighting coefficient of the movement trajectory to 0.3 and the weighting coefficient of the electromyography (EMG) signal to 0.7. If the patient's movement trajectory basically meets the standard, but the EMG signal intensity is insufficient, the deviation of the model calculation will increase significantly, and the real-time feedback unit will prompt to increase the force exerted.

[0113] If the patient also has mild shoulder joint adhesions, making it difficult to achieve the standard range of motion, but with good muscle strength recovery, medical staff can use a medical terminal to lower the weighting coefficient of the movement trajectory to 0.6 and raise the weighting coefficient of the electromyography signal to 0.4. After adjustment, the assessment results will place more emphasis on muscle strength performance, avoiding inflated biases caused by limited joint movement, and making the assessment more consistent with the patient's actual recovery status.

[0114] The dynamic weighting coefficient adjustment function solves the problem of a single assessment standard for different types of rehabilitation training by automatically adapting weights to ensure a precise match between assessment focus and training goals. The manual adjustment function takes into account individual patient differences, enhancing the flexibility and personalization of the assessment. The combination of these two adjustment methods makes the calculation of movement deviation more targeted and scientific, providing accurate assessments for patients with different physical conditions at different training stages. This ensures the effectiveness of real-time feedback and supplementary education, promoting the efficient implementation of rehabilitation training.

[0115] Optionally, the cloud-based data management device is used to store basic patient information, motion collection data, and education effectiveness evaluation results, build and update a risk scenario database and a rehabilitation training motion parameter database, and use big data analysis to identify weaknesses in education for different patient groups.

[0116] Specifically, the cloud-based data management device serves as the core of data storage and resource optimization support for the entire education and outreach system. By comprehensively storing various key data, maintaining and updating the core resource database, and uncovering patterns in group education and outreach, it provides a solid guarantee for personalized education and overall system optimization. Its various functions are interconnected, forming a complete closed loop of data collection, storage, analysis, and application.

[0117] The cloud-based data management device primarily handles comprehensive data storage. The stored content encompasses three core categories: patient basic information, motion capture data, and education effectiveness evaluation results. Patient basic information includes age, disease type, surgical procedure, and rehabilitation training stage, providing a foundation for scenario matching and personalized assessment. Motion capture data consists of raw data collected in real-time by the motion capture and data acquisition device during the education process, including motion trajectory data, electromyographic signal data, and 3D motion model data. This data is stored with timestamps linked to patient information for easy subsequent traceability and analysis. Education effectiveness evaluation results include assessment indicators such as scenario interaction accuracy, motion deviation, average question-and-answer accuracy, and mastery level, forming a personal education file for each patient and recording the entire process and effectiveness of rehabilitation education.

[0118] Meanwhile, the cloud-based data management device is responsible for building and continuously updating the risk scenario library and the rehabilitation training movement parameter library. The initial risk scenario library includes simulated sub-scenarios of common clinical risks such as falls, bed falls, and dislocations, with each scenario having pre-set complete interactive branches and scenario resources. The rehabilitation training movement parameter library stores standard movement data for different diseases and training types, including parameter thresholds such as movement trajectory vectors, standard electromyographic signal intensity, and movement angle ranges. During updates, the cloud-based data management device integrates new clinical cases and rehabilitation medicine research findings, supplements new risk scenarios and training parameters, and optimizes the interactive logic and parameter standards of existing scenarios to ensure the timeliness and professionalism of the resource library.

[0119] Furthermore, the cloud-based data management device possesses big data analytics capabilities, enabling the aggregation and analysis of massive amounts of stored patient data. The analysis process focuses on the characteristics of different patient groups, categorizing them by dimensions such as age, disease type, and rehabilitation stage, and identifying common behaviors and weaknesses across various groups during the education process. For example, it analyzes common errors in risky scenarios among elderly patients, or common motor deviations among stroke patients during specific rehabilitation training, summarizing patterns in group education.

[0120] Taking elderly post-orthopedic surgery patients as an example, the cloud-based data management device stores basic information, movement data collection, and assessment results for all patients in this group. Big data analysis revealed that in fall risk simulation scenarios, 60% of these patients made the mistake of getting out of bed without holding onto the handrail. In hip flexion and extension rehabilitation training, the average movement deviation was between 0.3 and 0.5, with weaknesses including balance control when getting up and insufficient hip flexion angle. Based on these findings, the cloud-based data management device updated its risk scenario database, added guidance on balance control when getting up, and optimized the standard movement prompts for this training in the rehabilitation training movement parameter database, making subsequent education for similar patients more targeted.

[0121] The comprehensive storage capabilities of the cloud-based data management system ensure the integrity and traceability of patient education data, providing a historical basis for adjusting individual education plans. Dynamic updates to the risk scenario database and rehabilitation training movement parameter database guarantee that the education content always aligns with clinical practice and medical standards. Data analytics uncovers group vulnerabilities, providing data support for optimizing the overall system's education content, extending from precise individual education to addressing common group problems. These functions work synergistically, enabling the education system to meet patients' individualized needs while continuously improving the overall quality and efficiency of education, providing scientific decision-making support for medical staff, and promoting the standardization and intelligent development of rehabilitation education.

[0122] The modules in the aforementioned patient education system based on augmented reality scenario simulation can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0123] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc. Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0124] Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A patient education method based on augmented reality scenario simulation, characterized in that, include: Medical staff enter patients’ basic information through medical terminals, select corresponding risk simulation scenarios and rehabilitation training scenarios from the scenario library of the cloud data management device, and push them to the augmented reality wearable terminal. The patient wears an augmented reality wearable terminal to enter a virtual scene. The motion capture and data acquisition device collects the patient's motion trajectory data and electromyographic signal data in real time, which are then transmitted to the intelligent assessment and feedback device and the virtual scene construction and interaction device, respectively. The virtual scene construction and interaction device drives the branch evolution of the virtual scene based on the received motion data; the intelligent assessment and feedback device calculates the motion deviation degree through the motion deviation analysis model; and the real-time feedback unit provides voice prompts and visual prompts to the patient through the augmented reality wearable terminal. After the education campaign concludes, the education effectiveness evaluation model generates an evaluation report and uploads it to the cloud data management device. Medical staff can view the evaluation report through their medical terminals and push personalized supplementary education content to augmented reality wearable terminals.

2. The patient education method based on augmented reality scenario simulation according to claim 1, characterized in that, The intelligent evaluation and feedback device achieves data fusion through a spatiotemporal synchronization calibration algorithm, including: using the action trigger time of the inertial measurement unit as the reference timestamp T0; detecting the signal trigger time T1 of the electromyography (EMG) sensor and calculating the time difference ΔT = T1 - T0; shifting the EMG signal along the time axis according to the difference between ΔT and ΔT0 to achieve time alignment between the EMG signal and the action trajectory data; and performing spatiotemporal correlation fusion between the aligned EMG signal and the action trajectory data; where ΔT0 is the EMG-action synchronization time difference of the standard action.

3. A patient education system based on augmented reality scenario simulation, characterized in that, include: Augmented reality wearable terminals, motion capture and data acquisition devices, virtual scene construction and interaction devices, intelligent assessment and feedback devices, cloud data management devices, and medical terminals; The devices exchange data via wireless communication protocols; Augmented reality wearable devices are used to present virtual scenes and receive voice commands from patients; The motion capture and data acquisition device is used to collect patient movement trajectory data and electromyographic signal data and transmit them to the intelligent assessment and feedback device and the virtual scene construction and interaction device. The virtual scene construction and interaction device is used to construct risk simulation sub-scenes and rehabilitation training sub-scenes, and drives the evolution of scene branches based on the received patient action data; The intelligent assessment and feedback device is used to analyze movement deviations and generate assessment results, providing feedback to patients through an augmented reality wearable terminal; The cloud-based data management device is used to store data and update the scene and parameter libraries; the medical terminal is used to view the evaluation results and push personalized educational content to the augmented reality wearable terminal.

4. A patient education system based on augmented reality scenario simulation according to claim 3, characterized in that, The augmented reality wearable terminal includes a high-definition display screen, a voice interaction unit, a positioning unit, a haptic feedback unit, and an environmental sound effect simulation unit; the high-definition display screen is used to present virtual scenes and virtual models for rehabilitation training; the voice interaction unit is used to receive voice commands from patients; and the positioning unit is used to obtain the physical spatial location information of patients. The haptic feedback unit is used to generate vibration feedback at the corresponding limb of the patient; the environmental sound effect simulation unit is used to play the environmental sound effects corresponding to the virtual scene.

5. A patient education system based on augmented reality scenario simulation according to claim 3, characterized in that, The motion capture and data acquisition device includes an inertial measurement unit, an electromyography (EMG) sensor, and a depth camera. The inertial measurement unit is used to acquire angular velocity and acceleration data of the patient's head posture and limb movements. The EMG sensor is used to acquire the EMG signal intensity of the target muscle groups for rehabilitation training. The depth camera is used to capture the contours of the patient's whole body movements and generate a three-dimensional motion model.

6. A patient education system based on augmented reality scene simulation according to claim 3, characterized in that, The virtual scene construction and interaction device includes a scene library and an interaction logic engine; the scene library contains risk simulation sub-scenes and rehabilitation training sub-scenes, the risk simulation sub-scenes preset correct behavior interaction branches and incorrect behavior interaction branches, and the rehabilitation training sub-scenes preset standard action parameter thresholds. The interaction logic engine is used to receive motion data transmitted by motion capture and data acquisition devices, determine whether the patient's actions conform to the preset logic of the scene, and drive the evolution of virtual scene branches.

7. A patient education system based on augmented reality scenario simulation according to claim 3, characterized in that, The intelligent assessment and feedback device includes a motion deviation analysis model, a publicity and education effectiveness assessment model, and a real-time feedback unit; The motion deviation analysis model is used to receive three-dimensional motion data, electromyographic signal data and standard motion parameter library data from patients, and calculates motion deviation degree through the motion deviation degree calculation formula. The education effectiveness evaluation model is used to generate a patient mastery level by combining the accuracy rate of scene interaction, the mean value of action deviation, and the accuracy rate of education questions and answers. The real-time feedback unit is used to provide patients with movement correction information through voice prompts and visual cues.

8. A patient education system based on augmented reality scene simulation according to claim 7, characterized in that, include: The formula for calculating the motion deviation is D=α×(1-cosθ)+β×(1-S p / S s ); where D is the movement deviation; α is the movement trajectory weighting coefficient; θ is the angle between the patient's movement trajectory vector and the standard movement trajectory vector; β is the electromyographic signal weighting coefficient; S p S represents the average actual electromyographic signal intensity of the patient's muscle groups. s This represents the average electromyographic signal intensity of the muscle group corresponding to the standard movement.

9. A patient education system based on augmented reality scene simulation according to claim 7, characterized in that, The motion deviation analysis model is equipped with a dynamic weight coefficient adjustment function, which adjusts the values ​​of α and β based on the type of rehabilitation training; when the rehabilitation training type is joint mobility training, the value of α is 0.8 and the value of β is 0.2; when the rehabilitation training type is muscle strength training, the value of α is 0.3 and the value of β is 0.

7. The weighting coefficients can be manually adjusted via the medical terminal.

10. A patient education system based on augmented reality scene simulation according to claim 3, characterized in that, The cloud-based data management device is used to store basic patient information, action data collection, and education effectiveness evaluation results. It also builds and updates a risk scenario database and a rehabilitation training action parameter database, and uses big data analysis to identify weaknesses in education for different patient groups.

Images