AI-assisted limb rehabilitation system

Abstract

The invention discloses an AI auxiliary limb rehabilitation system, and belongs to the field of communication control systems, the AI auxiliary limb rehabilitation system combines a sensing module, an AI analysis module, a VR interaction module, a feedback module and the like, comprehensively collects limb movement and physiological data, introduces gamification elements, improves the patient participation degree through target setting, an award mechanism and interesting tasks, and improves the rehabilitation effect. A personalized rehabilitation plan is generated by using a deep learning algorithm, an emotion recognition unit can adjust the plan according to a psychological state, environment self-adaption and real-time adjustment of training difficulty to ensure training safety and is challenging, data cloud storage facilitates remote monitoring, third-party equipment is compatible with extended monitoring ability, and the training efficiency is improved. A brand-new, efficient and personalized solution system scheme is provided for limb rehabilitation.

Description

Technical Field

[0001] The present invention relates to the field of communication control systems, and more specifically, to an AI-assisted limb rehabilitation system. Background Art

[0002] Limb rehabilitation plays a vital role in the modern medical field. Whether it is due to accidental injury or disease that causes limb function impairment, or patients who need to restore limb mobility after surgery, they urgently need effective rehabilitation methods to help them regain normal life and work ability.

[0003] Traditional limb rehabilitation methods mainly rely on manual guidance from doctors and simple rehabilitation equipment. In this model, the rehabilitation process often lacks personalized customization. Different patients have huge differences in their conditions, physical conditions and rehabilitation needs. It is difficult for traditional methods to formulate the most suitable rehabilitation plan for each patient's specific situation. At the same time, traditional rehabilitation methods are also relatively boring, lacking in fun and interactivity, and it is difficult to stimulate patients' active participation. Patients are prone to fatigue and boredom during the long rehabilitation process, which reduces their compliance with rehabilitation. In addition, the real-time feedback mechanism of traditional rehabilitation methods is relatively weak. When patients are undergoing rehabilitation training, it is difficult for them to accurately understand whether their movements are correct and whether they have met the requirements of rehabilitation. Doctors are also unable to monitor the patient's rehabilitation progress and movement performance in real time and can only adjust the rehabilitation plan through regular examinations and evaluations, which leads to inefficient rehabilitation process and may miss the best time for rehabilitation.

[0004] With the continuous advancement of science and technology, the rapid development of artificial intelligence and virtual reality technology has brought new opportunities for limb rehabilitation. Combining artificial intelligence with virtual reality technology to create a new AI-assisted limb rehabilitation system has become an urgent need in the current medical rehabilitation field. Summary of the invention

[0005] 1. Technical issues to be solved In response to the problems existing in the prior art, the purpose of the present invention is to provide an AI-assisted limb rehabilitation system, which combines a sensor module, an AI analysis module, a VR interaction module and a feedback module, etc., to comprehensively collect limb movement and physiological data, and introduce gamification elements. It improves patient participation through goal setting, reward mechanisms and interesting tasks, and uses deep learning algorithms to generate personalized rehabilitation plans. The emotion recognition unit can adjust the plan according to the psychological state. Environmental adaptation and real-time adjustment of training difficulty ensure that the training is safe and challenging. Data cloud storage facilitates remote monitoring, and third-party device compatibility expands monitoring capabilities, providing a new, efficient and personalized solution system for limb rehabilitation.

[0006] 2. Technical solution

[0007] To solve the above problems, the present invention adopts the following technical solutions.

[0008] AI-assisted limb rehabilitation system, including: A sensor module, used to obtain the user's body movement data; an AI analysis module, which is in communication with the sensor module, and is used to process and analyze the limb movement data, generate a user's rehabilitation assessment report based on a deep learning algorithm, and provide personalized suggestions for the user's rehabilitation progress; VR interaction module, which is used to provide users with virtual reality rehabilitation scenes, enhance users' rehabilitation training experience through tasks and guidance in the virtual environment, and enable users to complete specific rehabilitation movements in an immersive environment; A feedback module, communicating with the VR interaction module and the AI ​​analysis module, for providing real-time feedback on the user's body movements according to the AI ​​analysis results, prompting corrective movements, and enhancing training effects; A data cloud storage module is used to store the user's limb movement data and rehabilitation process data in the cloud so that doctors and rehabilitation experts can remotely monitor and guide the user's rehabilitation process. The data cloud storage module adopts safe and reliable cloud storage technology to ensure the privacy and security of user data. Doctors and rehabilitation experts can remotely access cloud data to understand the user's rehabilitation progress and status in real time and provide personalized guidance and suggestions to users. User incentive module, which encourages users to actively participate in rehabilitation training through gamification reward mechanism and social elements, so as to improve rehabilitation compliance and long-term training effect; The system can adjust the training difficulty in real time through the collection and motion data analysis module to ensure that the rehabilitation training is both challenging and adaptable to the patient's ability level. The system collects the patient's motion data through the sensor module, and the AI ​​analysis module analyzes this data to adjust the training difficulty in real time according to the patient's rehabilitation progress and ability. When the patient's ability improves, the system will automatically increase the training difficulty to keep the training challenging. When the patient becomes tired or has difficulty, the system will reduce the training difficulty to ensure that the patient can complete the training smoothly.

[0009] Furthermore, the sensing module is communicatively connected to a wearable sensor, a camera and a physiological signal acquisition device to capture the user's limb posture, motion trajectory and physiological parameters, wherein the wearable sensor adopts a high-precision inertial measurement unit, which can accurately detect the user's joint angle and motion acceleration; the camera has high resolution and high frame rate, and can capture the user's limb movements in real time; the physiological signal acquisition device is compatible with a variety of third-party wearable devices, expanding the monitoring capability of the user's physiological signals, including but not limited to heart rate, blood oxygen concentration, electromyographic signals and electroencephalographic signals, thereby providing more comprehensive data support for rehabilitation evaluation.

[0010] Furthermore, the AI ​​analysis module includes: The motion trajectory prediction unit is used to predict the user's limb motion trajectory through an advanced neural network model and compare the difference between the actual action and the target action. The motion trajectory prediction unit uses a combination of a long short-term memory network (LSTM) and a convolutional neural network (CNN) to accurately predict complex limb motion trajectories and provide real-time feedback. The specific calculation formula of the difference value is: difference value = ∑(|actual action parameter - target action parameter|); A rehabilitation assessment unit is used to assess the user's rehabilitation progress based on multi-dimensional limb data and physiological parameters, and to provide a dynamically adjusted rehabilitation plan in combination with historical data. The rehabilitation assessment unit adopts a multi-index comprehensive assessment method, including indicators such as joint range of motion, muscle strength, balance ability and endurance, and formulates a personalized rehabilitation plan for the user through data analysis and machine learning algorithms. The calculation formula of the multi-index comprehensive assessment method is: rehabilitation progress score = α×joint range of motion score + β×muscle strength score + γ×balance ability score + δ×endurance score, where α, β, γ, and δ are weight coefficients, which are adjusted according to different rehabilitation stages and patient conditions; The emotion recognition unit analyzes the user's facial expressions and voice emotions during the rehabilitation process, evaluates the user's psychological state, and dynamically adjusts the rehabilitation plan to increase the user's comfort and confidence. The emotion recognition unit uses computer vision and voice processing technology to monitor the user's emotional changes in real time. When the user has negative emotions such as anxiety and depression, the system will automatically adjust the rehabilitation plan to increase fun and interactivity and improve user participation.

[0011] Furthermore, the VR interaction module includes: A task generation unit is used to generate adaptive virtual training tasks according to the user's rehabilitation needs and abilities to improve the interest and participation of rehabilitation. The task generation unit automatically generates virtual training tasks of different difficulty levels according to the user's rehabilitation stage and goals, such as balance training, strength training, and coordination training. The difficulty of the task can be automatically adjusted according to the user's performance to ensure that the user continues to make progress in the challenge. The difficulty calculation formula of the task is: task difficulty coefficient = f (user performance parameter), where f is a functional relationship determined based on experience; The environment adaptation unit is used to automatically adjust the scene difficulty and task complexity in virtual reality according to the user's real-time feedback and physiological state to ensure that the training process is safe and challenging. The environment adaptation unit automatically adjusts the difficulty and complexity of the virtual environment by real-time monitoring of the user's physiological parameters and athletic performance. For example, when the user's heart rate is too high or there are signs of fatigue, the system will automatically reduce the task difficulty to ensure the safety of the training process. The calculation formula for the scene difficulty and task complexity is: scene difficulty adjustment value = g (physiological parameter change value, athletic performance change value), where g is a functional relationship determined based on experience.

[0012] Furthermore, the feedback module includes: A real-time voice feedback unit is used to provide voice prompts when the user completes an action to help the user correct the wrong action. The real-time voice feedback unit uses natural language processing technology to provide personalized voice prompts based on the user's actions and performance, such as "raise your arms a little higher" or "take bigger steps", etc. The visual feedback unit is used to display the correctness of the user's actions through animation or marking in a virtual reality environment to help the user understand how to improve the quality of the action. The visual feedback unit uses virtual reality technology to display the user's action trajectory and correctness in a virtual environment. For example, it uses color markings to show whether the user's joint angles are correct, and demonstrates the correct action method through animation.

[0013] Furthermore, the user incentive module includes: The synchronous game unit introduces gamification elements to improve the patient's participation and enthusiasm by setting goals, reward mechanisms and interesting rehabilitation tasks. For example, different levels of rehabilitation tasks are set, and patients can get corresponding rewards after completing a level of tasks, such as virtual props, points, etc. At the same time, the system can set rehabilitation scenes with different themes according to the patient's preferences to increase the fun of rehabilitation training; Achievement system unit: when users complete specific rehabilitation tasks or reach certain rehabilitation goals, they will receive corresponding achievements and rewards, such as medals and trophies. The achievement system unit can stimulate users’ sense of competition and achievement, and improve their participation. The social interaction platform unit allows users to communicate and share with other rehabilitation patients, encourage and support each other. The social interaction platform unit can increase users' social connections and improve their rehabilitation compliance.

[0014] Furthermore, the feedback module is also communicatively connected to a safety protection module for monitoring the safety status of the user during rehabilitation training, and issuing an alarm in time and stopping training when an abnormal situation occurs. The safety protection module includes: The fall detection unit detects whether the user falls through sensors and algorithms. When a fall is detected, an alarm is immediately issued and relevant personnel are notified. The fall detection unit uses sensors such as acceleration sensors and gyroscopes, combined with machine learning algorithms, to accurately detect the user's fall and issue an alarm in time. The fall detection calculation formula is: fall probability = l (sensor acceleration change value, gyroscope change value), where l is a functional relationship determined based on experience; The motion limit monitoring unit is used to monitor the user's motion limit. When the user's motion exceeds the safe range, an alarm is issued in time and the training plan is adjusted. The motion limit monitoring unit monitors the user's joint angle, motion speed, strength and other parameters through sensors and algorithms. When the user's motion exceeds the safe range, the system will automatically issue an alarm and adjust the training plan to ensure the user's safety. The calculation formula of the motion limit monitoring unit is: probability of exceeding the safety range = m (joint angle change value, motion speed change value, strength change value), where m is a functional relationship determined based on experience.

[0015] 3. Beneficial effects

[0016] Compared with the prior art, the advantages of the present invention are: (1) This solution first connects wearable sensors, cameras, and physiological signal acquisition devices on the sensor module to capture the user's limb posture, movement trajectory, and various physiological parameters, including heart rate, blood oxygen concentration, electromyographic signals, and electroencephalographic signals, to provide comprehensive data support for rehabilitation evaluation. Then, based on the deep learning algorithm, the AI ​​analysis module can evaluate the user's rehabilitation progress based on multi-dimensional limb data and physiological parameters, and provide a dynamically adjusted rehabilitation plan based on historical data to meet the specific needs of different users and achieve personalized rehabilitation plan effects. (2) In this scheme, the synchronous game unit and achievement system unit set in the user incentive module not only introduce gamification elements into the training process, but also improve the patient's participation and enthusiasm by setting goals, reward mechanisms and interesting rehabilitation tasks. It also enables users to obtain corresponding achievements and rewards when completing specific rehabilitation tasks or achieving certain rehabilitation goals, which stimulates users' sense of competition and achievement and improves their participation. At the same time, the setting of the social interaction platform unit allows users to communicate and share with other rehabilitation patients, encourage and support each other. The social interaction platform unit can increase users' social connections and improve their rehabilitation compliance. (3) In this solution, the emotion recognition unit in the AI ​​analysis module evaluates the user's psychological state by analyzing the user's facial expressions and voice emotions during the rehabilitation process, and dynamically adjusts the rehabilitation plan to increase the user's comfort and confidence; (4) In this solution, the environmental adaptive unit of the VR interaction module automatically adjusts the scene difficulty and task complexity in the virtual reality according to the user's real-time feedback and physiological state, ensuring that the training process is safe and challenging. In addition, the safety protection module connected to the feedback module can monitor the user's safety status during rehabilitation training in real time. When an abnormal situation occurs, an alarm is issued in time and the training is stopped, thereby improving the safety of the training. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic diagram of the overall system architecture of the present invention; Figure 2 Schematic diagram of the subunit composition of the AI ​​analysis module of the present invention; Figure 3 It is a schematic diagram of the sub-unit composition of the VR interaction module of the present invention; Figure 4 Schematic diagram of the subunit composition of the feedback module of the present invention; Figure 5 It is a schematic diagram of the sub-unit composition of the user incentive module of the present invention; Figure 6 It is a schematic diagram of the sub-unit composition of the security protection module of the present invention. DETAILED DESCRIPTION

[0018] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the specification of the present invention; it is obvious that the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments, and all other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making creative work are within the scope of protection of the present invention.

[0019] Embodiment 1:

[0020] See also Figure 1-Figure 6 When a patient suffers from hemiplegia of limbs due to disease or other factors and his ability to take care of himself is affected, he can start using the AI-assisted limb rehabilitation system of the present invention for rehabilitation training under the recommendation of a doctor, as follows: First, under the guidance of professionals, the patient correctly wears wearable sensors, cameras, and physiological signal acquisition devices. These devices fit the patient's body tightly and can accurately capture the patient's limb posture, movement trajectory, and physiological parameters. Among them, the high-precision inertial measurement unit in the wearable sensor is like a sensitive antenna, constantly sensing the subtle changes in the patient's joint angles and fluctuations in movement acceleration. The high-resolution, high-frame-rate camera is like a loyal observer, recording the patient's limb movements in all directions without missing any details. The physiological signal acquisition device can not only monitor common physiological parameters such as heart rate and blood oxygen concentration, but is also compatible with a variety of third-party wearable devices, further expanding the monitoring capabilities of patients' electromyographic signals and electroencephalographic signals, and providing extremely comprehensive data support for rehabilitation evaluation; Then, the AI ​​analysis module quickly processes and analyzes the collected data. The motion trajectory prediction unit uses an advanced neural network model, like a wise prophet, to accurately predict the patient's limb motion trajectory. It carefully compares the actual movement with the target movement and calculates the difference value = ∑(|actual movement parameter - target movement parameter|), which provides a clear direction for subsequent rehabilitation guidance. The rehabilitation assessment unit is like a rigorous referee. According to multi-dimensional limb data and physiological parameters, it uses a multi-index comprehensive evaluation method such as joint range of motion, muscle strength, balance ability and endurance to score the patient's rehabilitation progress. The rehabilitation progress score = α× joint range of motion score + β× muscle strength score + γ× balance ability score + δ× endurance score, where α, β, γ, δ are weight coefficients, which are adjusted according to different rehabilitation stages and patient conditions. The emotion recognition unit is also silently paying attention to the patient's psychological state. By analyzing the patient's facial expressions and voice emotions during the rehabilitation process, it is like a caring friend. When the patient has negative emotions, it adjusts the rehabilitation plan in time to increase fun and interactivity and give the patient warm support. Then the VR interaction module provides a virtual reality rehabilitation scene, making the patient feel as if he is in a whole new world. The task generation unit generates adaptive virtual training tasks based on the patient's rehabilitation needs and abilities. These tasks are like interesting challenges that inspire the patient's fighting spirit. For example, in the balance training task, the patient needs to maintain body balance on the virtual shaking platform. As the patient's performance improves, the task difficulty coefficient = f (user performance parameter) will also be adjusted accordingly, so that the patient can always stay in a state of challenge and growth. The environmental adaptive unit is like an intelligent guardian. According to the patient's real-time feedback and physiological state, it automatically adjusts the scene difficulty and task complexity in virtual reality. When the patient's heart rate increases or signs of fatigue appear, it will respond quickly to reduce the difficulty of the task and ensure the safety of the training process. The calculation formula is: scene difficulty adjustment value = g (physiological parameter change value, sports performance change value), where g is a functional relationship determined based on experience; At the same time, the feedback module is also playing an important role at all times. The real-time voice feedback unit provides caring voice prompts when the patient completes the action, "Raise your arms a little higher, very good!" "Take more steady steps." These personalized voice prompts help patients correct wrong actions, and the visual feedback unit shows the correctness of the user's actions through animation or markings in a virtual reality environment. When the patient's joint angle is correct, a green mark will appear; when the action needs to be improved, there will be an animation demonstrating the correct action method, so that the patient can see it at a glance; Then the data cloud storage module stores the patient's limb movement data and rehabilitation process data securely in the cloud. Doctors and rehabilitation experts can remotely monitor and guide the patient's rehabilitation process and adjust the rehabilitation plan at any time according to the patient's condition. The achievement system unit in the user incentive module sets different levels of rehabilitation tasks for patients. When patients complete a level of task, they will receive corresponding rewards, such as virtual medals, trophies, etc., which greatly improves the patient's participation and enthusiasm. The participation score = h (rationality of goal setting, attractiveness of reward mechanism, and fun of task), where h is a functional relationship determined based on experience. The social interaction platform unit also gives patients the opportunity to communicate and share experiences with other rehabilitation patients, encourage and support each other, and increase patients' social connections and rehabilitation compliance; Rehabilitation training is carried out continuously in this way until the patient's limb function is significantly improved and recovered.

[0021] Embodiment 2:

[0022] See also Figure 1-Figure 6 When an athlete suffers from knee injury due to injury and needs to use the AI-assisted limb rehabilitation system of the present invention for rehabilitation training: First, the athletes put on wearable sensors, cameras, and physiological signal acquisition devices. The wearable sensors accurately detect the athlete's joint angles and movement acceleration, providing an accurate data basis for subsequent analysis. The camera records the athlete's body movements at high resolution and high frame rate. The physiological signal acquisition device not only monitors conventional physiological parameters, but also obtains more comprehensive heart rate, blood oxygen concentration, electromyographic signal and other data through compatibility with third-party wearable devices, providing strong support for rehabilitation evaluation. Then the motion trajectory prediction unit in the AI ​​analysis module uses a combination of long short-term memory network (LSTM) and convolutional neural network (CNN) to accurately predict the athlete's limb motion trajectory. It compares the actual movement with the target movement and provides real-time feedback to the athlete. The rehabilitation assessment unit evaluates the athlete's rehabilitation progress based on multi-dimensional limb data and physiological parameters. The rehabilitation progress score = α× joint range of motion score + β× muscle strength score + γ× balance ability score + δ× endurance score, where α, β, γ, and δ are weight coefficients, which are dynamically adjusted according to the athlete's rehabilitation stage and physical condition. The emotion recognition unit also pays attention to the athlete's psychological state to ensure that he maintains a positive attitude during the rehabilitation process. In the VR interaction module, the task generation unit generates a series of challenging virtual training tasks based on the athletes' rehabilitation needs and abilities. For example, in the strength training task, the athlete needs to lift a barbell that is gradually getting heavier in the virtual environment. As the athlete's ability improves, the task difficulty coefficient = f (user performance parameter) will continue to increase, allowing the athlete to always remain in a state of challenging himself. The environmental adaptive unit automatically adjusts the scene difficulty and task complexity in virtual reality based on the athlete's real-time feedback and physiological state. When the athlete's heart rate is too high or there are signs of fatigue, it will automatically reduce the task difficulty to ensure the safety of the training process. The calculation formula is: scene difficulty adjustment value = g (physiological parameter change value, sports performance change value), where g is a functional relationship determined based on experience; The feedback module provides timely and effective feedback to athletes. The real-time voice feedback unit provides professional voice prompts during the athletes' movements, such as "Make the force point more accurate!" and "Make the movement smoother." This helps athletes correct their incorrect movements. The visual feedback unit uses animations or marks to show the correctness of the athletes' movements in a virtual reality environment. When the athletes' movements are standard, a golden mark will appear; when the movements need to be improved, there will be animations demonstrating the correct movement methods, allowing athletes to quickly adjust. The data cloud storage module stores the athlete's limb movement data and rehabilitation process data in the cloud. Coaches and rehabilitation experts can remotely monitor and guide the athlete's rehabilitation process. The achievement system unit in the user incentive module sets higher rehabilitation goals for athletes. When an athlete completes a goal, he will receive corresponding rewards and honors, which motivates him to constantly challenge himself. The social interaction platform unit also gives athletes the opportunity to exchange rehabilitation experiences with other athletes, and encourage and support each other. Continue rehabilitation training in this way until the athlete's limb function is significantly improved and recovered.

[0023] The above is only a preferred specific implementation of the present invention; however, the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical solution and its improved conception of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. AI-assisted limb rehabilitation system, characterized in that: include: A sensor module, used to obtain the user's body movement data; An AI analysis module, which is in communication with the sensor module, is used to process and analyze the limb movement data, generate a user's rehabilitation assessment report based on a deep learning algorithm, and provide personalized suggestions for the user's rehabilitation progress; VR interaction module is used to provide users with virtual reality rehabilitation scenes, enhance users' rehabilitation training experience through tasks and guidance in the virtual environment, and enable users to complete specific rehabilitation movements in an immersive environment; A feedback module, communicating with the VR interaction module and the AI ​​analysis module, for providing real-time feedback on the user's body movements according to the AI ​​analysis results, prompting corrective movements, and enhancing training effects; A data cloud storage module is used to store the user's limb movement data and rehabilitation process data in the cloud so that doctors and rehabilitation experts can remotely monitor and guide the user's rehabilitation process. The data cloud storage module adopts safe and reliable cloud storage technology to ensure the privacy and security of user data. Doctors and rehabilitation experts can remotely access cloud data to understand the user's rehabilitation progress and status in real time and provide personalized guidance and suggestions to users. A user incentive module, which is in communication with the feedback module and motivates users to actively participate in rehabilitation training through a gamification reward mechanism and social elements, so as to improve rehabilitation compliance and long-term training effects; The system can adjust the training difficulty in real time through the collection and motion data analysis module to ensure that the rehabilitation training is both challenging and adaptable to the patient's ability level. The system collects the patient's motion data through the sensor module, and the AI ​​analysis module analyzes this data to adjust the training difficulty in real time according to the patient's rehabilitation progress and ability. When the patient's ability improves, the system will automatically increase the training difficulty to maintain the challenging nature of the training. When the patient becomes tired or has difficulty, the system will reduce the training difficulty to ensure that the patient can complete the training smoothly.

2. The AI-assisted limb rehabilitation system according to claim 1, characterized in that: The sensing module is communicatively connected with a wearable sensor, a camera and a physiological signal acquisition device to capture the user's limb posture, motion trajectory and physiological parameters. The wearable sensor adopts a high-precision inertial measurement unit, which can accurately detect the user's joint angle and motion acceleration. The camera has high resolution and high frame rate, and can capture the user's limb movements in real time. The physiological signal acquisition device is compatible with a variety of third-party wearable devices, expanding the monitoring capability of the user's physiological signals, including but not limited to heart rate, blood oxygen concentration, electromyographic signals and electroencephalographic signals, thereby providing more comprehensive data support for rehabilitation evaluation.

3. The AI-assisted limb rehabilitation system according to claim 1, characterized in that: The AI ​​analysis module includes: The motion trajectory prediction unit is used to predict the user's limb motion trajectory through an advanced neural network model and compare the difference between the actual action and the target action. The motion trajectory prediction unit uses a combination of a long short-term memory network (LSTM) and a convolutional neural network (CNN) to accurately predict complex limb motion trajectories and provide real-time feedback. The specific calculation formula of the difference value is: difference value = ∑(|actual action parameter - target action parameter|); A rehabilitation assessment unit is used to assess the user's rehabilitation progress based on multi-dimensional limb data and physiological parameters, and to provide a dynamically adjusted rehabilitation plan in combination with historical data. The rehabilitation assessment unit adopts a multi-index comprehensive assessment method, including indicators such as joint range of motion, muscle strength, balance ability and endurance, and formulates a personalized rehabilitation plan for the user through data analysis and machine learning algorithms. The calculation formula of the multi-index comprehensive assessment method is: rehabilitation progress score = α×joint range of motion score + β×muscle strength score + γ×balance ability score + δ×endurance score, where α, β, γ, and δ are weight coefficients, which are adjusted according to different rehabilitation stages and patient conditions; The emotion recognition unit analyzes the user's facial expressions and voice emotions during the rehabilitation process, evaluates the user's psychological state, and dynamically adjusts the rehabilitation plan to increase the user's comfort and confidence. The emotion recognition unit uses computer vision and voice processing technology to monitor the user's emotional changes in real time. When the user has negative emotions such as anxiety and depression, the system will automatically adjust the rehabilitation plan to increase fun and interactivity and improve user participation.

4. The AI-assisted limb rehabilitation system according to claim 1, characterized in that: The VR interaction module includes: A task generation unit is used to generate adaptive virtual training tasks according to the rehabilitation needs and abilities of the user, so as to improve the interest and participation of rehabilitation. The task generation unit automatically generates virtual training tasks of different difficulty levels according to the rehabilitation stage and goals of the user, such as balance training, strength training and coordination training. The difficulty of the task can be automatically adjusted according to the performance of the user to ensure that the user continues to make progress in the challenge. The difficulty calculation formula of the task is: task difficulty coefficient = f (user performance parameter), where f is a functional relationship determined based on experience; The environment adaptation unit is used to automatically adjust the scene difficulty and task complexity in virtual reality according to the user's real-time feedback and physiological state to ensure that the training process is safe and challenging. The environment adaptation unit automatically adjusts the difficulty and complexity of the virtual environment by real-time monitoring of the user's physiological parameters and athletic performance. For example, when the user's heart rate is too high or there are signs of fatigue, the system will automatically reduce the task difficulty to ensure the safety of the training process. The calculation formula for the scene difficulty and task complexity is: scene difficulty adjustment value = g (physiological parameter change value, athletic performance change value), where g is a functional relationship determined based on experience.

5. The AI-assisted limb rehabilitation system according to claim 1, characterized in that: The feedback module comprises: A real-time voice feedback unit is used to provide voice prompts when the user completes an action to help the user correct wrong actions. The real-time voice feedback unit uses natural language processing technology to provide personalized voice prompts based on the user's actions and performance, such as "raise your arms a little higher" and "take bigger steps", etc.; The visual feedback unit is used to display the correctness of the user's actions through animation or marking in a virtual reality environment to help the user understand how to improve the quality of the action. The visual feedback unit uses virtual reality technology to display the user's action trajectory and correctness in a virtual environment. For example, it uses color markings to show whether the user's joint angles are correct, and demonstrates the correct action method through animation.

6. The AI-assisted limb rehabilitation system according to claim 1, characterized in that: The user incentive module includes: The synchronous game unit introduces gamification elements to improve the patient's participation and enthusiasm by setting goals, reward mechanisms and interesting rehabilitation tasks. For example, different levels of rehabilitation tasks are set, and patients can get corresponding rewards after completing a level of tasks, such as virtual props, points, etc. At the same time, the system can set rehabilitation scenes with different themes according to the patient's preferences to increase the fun of rehabilitation training; Achievement system unit: when users complete specific rehabilitation tasks or reach certain rehabilitation goals, they will receive corresponding achievements and rewards, such as medals and trophies. The achievement system unit can stimulate users’ sense of competition and achievement, and improve their participation. The social interaction platform unit allows users to communicate and share with other rehabilitation patients, encourage and support each other. The social interaction platform unit can increase users' social connections and improve their rehabilitation compliance.

7. The AI-assisted limb rehabilitation system according to claim 1, characterized in that: The feedback module is also communicatively connected to a safety protection module for monitoring the safety status of the user during rehabilitation training, and issuing an alarm and stopping training in a timely manner when an abnormal situation occurs. The safety protection module includes: The fall detection unit detects whether the user falls through sensors and algorithms. When a fall is detected, an alarm is immediately issued and relevant personnel are notified. The fall detection unit uses sensors such as acceleration sensors and gyroscopes, combined with machine learning algorithms, to accurately detect the user's fall and issue an alarm in time. The fall detection calculation formula is: fall probability = l (sensor acceleration change value, gyroscope change value), where l is a functional relationship determined based on experience; The motion limit monitoring unit is used to monitor the user's motion limit. When the user's motion exceeds the safe range, an alarm is issued in time and the training plan is adjusted. The motion limit monitoring unit monitors the user's joint angle, motion speed, strength and other parameters through sensors and algorithms. When the user's motion exceeds the safe range, the system will automatically issue an alarm and adjust the training plan to ensure the user's safety. The calculation formula of the motion limit monitoring unit is: probability of exceeding the safety range = m (joint angle change value, motion speed change value, strength change value), where m is a functional relationship determined based on experience.

Images