Integrated rehabilitation training system based on motion capture and myoelectricity feedback

Abstract

The invention relates to the technical field of medical data processing, in particular to an integrated rehabilitation training system based on motion capture and myoelectricity feedback, which comprises a motion data acquisition module, a motion evaluation and analysis module and a control module, and is specifically integrated on a training garment through inertial motion capture equipment and a surface myoelectricity sensor. The motion state and muscle activation condition of a patient are monitored in real time, actual motion data and actual electromyographic signal data of the patient are collected, and whether the motion state of the patient meets the standard or not is judged by comparing the data deviation condition between the actual motion of the patient and the standard motion. And according to the deviation analysis result, the electrical stimulation parameters can be dynamically adjusted through closed-loop control, so that personalized and precise spinal cord electrical stimulation rehabilitation training is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of medical data processing, and in particular to an integrated rehabilitation training system based on motion capture and electromyographic feedback. Background Art

[0002] In the field of rehabilitation of patients with spinal cord injury, existing rehabilitation training methods mostly rely on technologies such as functional electrical stimulation (FES) and epidural electrical stimulation (EES). These technologies help patients perform muscle activation and motor training by applying electrical stimulation to the spinal cord or specific muscle groups to promote neural remodeling and functional recovery. However, existing technologies have certain limitations in clinical applications.

[0003] Most existing rehabilitation systems lack a complete closed-loop feedback mechanism, resulting in poor synchronization between electrical stimulation and patient movement. Even though some systems can adjust electrical stimulation through simple sensor feedback, most of them are discrete or delayed adjustments, making it difficult to provide high-precision real-time feedback. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide an integrated rehabilitation training system based on motion capture and electromyographic feedback. The technical solutions adopted are as follows: The motion data acquisition module is used to collect the actual motion data and actual electromyographic signal data of the patient in the motion state, and obtain the standard motion data and standard electromyographic signal data of the standard motion in the rehabilitation training database; A motion evaluation and analysis module, used to identify the motion state according to the actual motion data and actual electromyographic signal data of the patient, and determine the actual motion of the patient; According to the difference between the actual motion data of the actual action of the patient and the standard motion data of the standard action, and the difference between the actual electromyographic signal data of the actual action and the standard electromyographic signal data of the standard action, a comprehensive deviation score and an actual deviation result of the patient are obtained; and whether the movement state of the patient meets the standard is judged according to the comprehensive deviation score; The control module is used to adjust the control parameters of the electrical stimulation according to the actual deviation results of the patient when the movement state of the patient does not meet the standard.

[0005] Preferably, the collecting of actual motion data of the patient in motion state specifically includes: The inertial detection unit in the training suit collects the three-axis acceleration and three-axis gyroscope data of each joint of the patient; the sampling frequency is at least 100Hz; Based on the three-axis acceleration and three-axis gyroscope data of various parts of the patient, the posture angle of the corresponding joint is calculated to obtain the joint angle and angular velocity of various parts of the patient; the patient's actual motion data includes the joint angle and angular velocity of various parts of the patient.

[0006] Preferably, the method for obtaining the patient's comprehensive deviation score and actual deviation result specifically includes: Based on the absolute value of the difference between the actual action of the patient and the standard action of the patient at the same joint position, a joint angle deviation value of the patient is obtained; Obtaining a muscle activation difference value of the patient based on an absolute value of a difference between a mean square error of the actual electromyographic signal data of the patient and a mean square error of the standard electromyographic signal data; The joint angle deviation value and the muscle activation difference value are weightedly summed to obtain a comprehensive deviation score for the patient; the actual deviation result includes the joint angle deviation value and the muscle activation difference value.

[0007] Preferably, adjusting the control parameters of the electrical stimulation according to the actual deviation result of the patient specifically includes: The PID controller is used to adjust the frequency, pulse width and amplitude of the electrical stimulation based on the actual deviation results of the patient to determine the control parameters.

[0008] Preferably, the output formula of the PID controller is: in, represents the output signal of the PID controller, Indicates the joint angle deviation value or muscle activation difference value, , and Represent the proportional, integral and differential gain coefficients respectively.

[0009] Preferably, judging whether the patient's movement state meets the standard according to the comprehensive deviation score specifically includes: When the comprehensive deviation score is greater than or equal to the preset evaluation threshold, the patient's motor status does not meet the standard; When the comprehensive deviation score is less than the preset evaluation threshold, the patient's movement status meets the standard.

[0010] Preferably, the system further comprises a user interaction adjustment module, which is used for patients and rehabilitation therapists to view motion assessment results and electromyographic feedback data through a user interface, and is also used for patients and rehabilitation therapists to manually adjust control parameters of electrical stimulation.

[0011] The embodiments of the present invention have at least the following beneficial effects: The present invention uses an integrated inertial motion capture device and surface electromyography sensor to capture the patient's joint posture and muscle activation status in real time. The control module uses an adaptive control algorithm and a closed-loop feedback mechanism to adjust the electrical stimulation parameters according to the patient's real-time data, thereby ensuring that the electrical stimulation is synchronized with the patient's motion state, improving the accuracy of the electrical stimulation and the rehabilitation effect.

[0012] By analyzing the patient's real-time movements and electromyographic feedback data, a personalized electrical stimulation program can be provided to meet the needs of patients at different stages of rehabilitation. Compared with traditional fixed-parameter electrical stimulation, the personalized adjustment mode of the present invention can greatly improve the pertinence and effect of training, shorten the rehabilitation cycle, and enhance the patient's recovery ability.

[0013] By fusing inertial and electromyographic data, the accuracy of motion capture and the reliability of data are improved, avoiding the problem of single sensor data being susceptible to interference. Through advanced data fusion algorithms, the system can more stably obtain high-precision patient motion data, providing solid data support for closed-loop control. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0015] Figure 1 It is a structural schematic diagram of an integrated rehabilitation training system based on motion capture and electromyographic feedback provided by the present invention; Figure 2 The closed-loop control flow chart provided by the present invention; Figure 3 A schematic diagram of a user interface provided by the present invention. DETAILED DESCRIPTION

[0016] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation method, structure, features and effects of an integrated rehabilitation training system based on motion capture and electromyographic feedback proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0017] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0018] The following is a detailed description of a specific scheme of an integrated rehabilitation training system based on motion capture and electromyographic feedback provided by the present invention in conjunction with the accompanying drawings.

[0019] See also Figure 1 , which shows a system structure diagram of an integrated rehabilitation training system based on motion capture and electromyographic feedback provided by an embodiment of the present invention. The system includes a motion data acquisition module, a motion evaluation and analysis module, a control module, and a user interaction adjustment module.

[0020] The main purpose of the present invention is to provide an integrated spinal cord electrical stimulation rehabilitation training system based on inertial motion capture and electromyographic feedback, which can realize closed-loop control by real-time monitoring of the patient's movement state and muscle activation, and dynamically adjusting the electrical stimulation parameters, thereby improving the rehabilitation effect and personalized training level of patients with spinal cord injury. The following is an introduction to the various modules of the system in conjunction with a specific embodiment.

[0021] The motion data acquisition module is used to collect the actual motion data and actual electromyographic signal data of the patient in the motion state, and obtain the standard motion data and standard electromyographic signal data of the standard motion in the rehabilitation training database.

[0022] First, it is necessary to obtain the standard motion trajectory and corresponding electromyographic signal pattern of the rehabilitation training action from the rehabilitation training database, for example, to obtain the standard motion trajectory data of walking training. The rehabilitation training database contains multiple typical motion trajectories and corresponding electrical stimulation parameters as the basis for comparison and training. That is, to obtain the standard motion data and standard electromyographic signal data of each standard action.

[0023] Before the training begins, the system records the patient's initial movement state through inertial motion capture and surface electromyography. The system compares the patient's initial posture data with the standard movement, determines the initial electrical stimulation parameters, and lays the foundation for real-time adjustment during the training process. That is, the patient's walking movements are captured in real time through the inertial motion capture device and surface electromyography sensor integrated in the training suit. The captured image data is used for posture recognition using a deep learning model to extract the three-dimensional coordinates and related kinematic data of the hip, knee and ankle joints, including joint angles, angular velocities, angular accelerations, gait cycle data, step frequency and step length, etc. The muscle activation of key parts such as the thigh, knee, and calf is obtained through electromyography data to evaluate the muscle activation sequence and coordination during walking.

[0024] Specifically, the inertial motion capture device and electromyographic sensor on the training suit can capture the patient's motion data and electromyographic feedback in real time, generate current posture data and muscle activation data, and monitor the patient's motion state, including data such as joint angle, speed and acceleration. The module is integrated in the key joints of the training suit to achieve real-time capture of the patient's motion posture. In this embodiment, the joints mainly include: shoulder joint, elbow joint, wrist joint, hip joint, knee joint, ankle joint.

[0025] Specifically, the three-axis acceleration and three-axis gyroscope data of each joint are collected in real time through multiple inertial measurement units (IMUs) embedded in the training suit. The sampling frequency is set above 100Hz to ensure the real-time and accuracy of motion capture. Furthermore, the posture angle of the corresponding joint is calculated based on the three-axis acceleration and three-axis gyroscope data of each part of the patient to obtain the joint angle and angular velocity of each part of the patient.

[0026] In this embodiment, a quaternion attitude solution algorithm is used to calculate the attitude angle of the joint through the acceleration and gyroscope data of the IMU sensor. And in the quaternion solution process, combined with the Kalman filter, the attitude data is subjected to real-time error correction and noise suppression to ensure the smoothness and stability of the joint attitude angle. After obtaining the attitude angle, the angular velocity of the joint is calculated using a numerical differentiation algorithm. To ensure the accuracy of the angular velocity calculation, a sliding window smoothing process is added to reduce the impact of instantaneous jitter on the calculation. The calculated attitude angle and angular velocity are used as the current kinematic features, and the data is sent to the control module for further analysis and motion evaluation.

[0027] At this point, the actual motion data of the patient can be obtained, and the actual motion data includes the joint angles and angular velocities of various parts of the patient. It can be understood that the standard motion data corresponding to the standard action also includes the joint angles and angular velocities of various parts, and each part refers to the joint part.

[0028] Furthermore, the surface electromyographic sensor array on the training suit is used to collect the electromyographic signals of the target muscle groups in real time. Each electromyographic sensor is provided with a sampling frequency of at least 500 Hz to accurately capture the rapidly changing signals of the muscles. Among them, the target muscle groups mainly include the upper limb muscle groups and the lower limb muscle groups. Specifically, the surface electromyography of the deltoid, levator scapulae, biceps, triceps, wrist flexors, wrist extensors, hip flexors, hip extensors, quadriceps, biceps femoris, semitendinosus, semimembranosus, gastrocnemius, soleus, and tibialis anterior is recorded. The electrophysiological signals (EMG) of the muscles are collected in real time through the electromyographic sensors to reflect the activation state and contraction strength of the muscles, and provide feedback for electrical stimulation regulation.

[0029] Then, the collected electromyographic signals are preprocessed and feature extracted. Specifically, the preprocessing operations of the electromyographic signals include bandpass filtering and normalization and baseline correction. The bandpass filtering operation specifically includes bandpass filtering (e.g., 20Hz-450Hz) of the original electromyographic signals to remove low-frequency noise (e.g., motion artifacts) and high-frequency interference (e.g., environmental electrical noise). The normalization and baseline correction operations specifically include normalizing the filtered signals to adapt to the electromyographic characteristics of different patients; at the same time, correcting the baseline drift to ensure the stability of the electromyographic data. For example, the surface electromyographic sensors on the training clothes monitor the activation of key muscle groups such as the hip flexors, hip extensors, quadriceps femoris and triceps surae in real time during walking, and record the RMS value and spectral characteristics of muscle activity in order to evaluate the muscle activation effect.

[0030] The feature extraction operation of electromyographic signals mainly includes time domain feature extraction and frequency domain feature extraction. Among them, time domain feature extraction specifically includes extracting indicators such as mean absolute value (MAV), root mean square value (RMS) and zero crossing point (ZC) in the time domain to quantify the intensity and frequency characteristics of muscle activation. Frequency domain feature extraction mainly includes extracting the main frequency component and spectrum power of electromyographic signals through fast Fourier transform (FFT) to judge the contraction and fatigue state of muscles.

[0031] At this point, the actual electromyographic signal data of the patient in the actual motion state can be collected.

[0032] It should be noted that, since the sampling frequencies of motion data and electromyographic signal data are different, further time synchronization processing is required. In this embodiment, a linear interpolation algorithm is used to align data of different frequencies to obtain consistent time series data. In other embodiments, the implementer can also obtain consistent time series data by setting the same sampling frequency.

[0033] Finally, a weighted average filtering algorithm is used to fuse the posture data and electromyographic feature data into a joint feature vector so that the control module can comprehensively analyze the movement and muscle activation status to enhance real-time performance and robustness.

[0034] The motion assessment and analysis module is used to identify the motion state according to the actual motion data and actual electromyographic signal data of the patient, and determine the actual motion of the patient; obtain the patient's comprehensive deviation score and actual deviation result according to the difference between the actual motion data of the patient's actual motion and the standard motion data of the standard motion, and the difference between the actual electromyographic signal data of the actual motion and the standard electromyographic signal data of the standard motion; and judge whether the patient's motion state meets the standard according to the comprehensive deviation score.

[0035] First, the fused data is classified in real time through the trained deep learning model to identify the current motion state. The motion state classification model inputs the joint feature vector and outputs the current motion state, that is, the actual motion of the patient, to provide real-time feedback data for the closed-loop control module. Among them, the deep learning model can be an LSTM model or a GRU network. In other embodiments, the current motion of the patient can also be identified by using the deep learning model through the collected motion trajectory of the patient. This method is a common known technology in the field and will not be introduced in detail here.

[0036] Then, the current motion state is compared with the standard state in the database and the deviation value is calculated. The deviation value is used to guide the closed-loop control module to adjust the electrical stimulation parameters to ensure that the patient's movements gradually approach the standard. In other words, the patient's current motion state is compared with the motion data of the standard movement, such as joint angle and electromyographic amplitude.

[0037] Specifically, in the first aspect, the joint angle deviation value of the patient is obtained based on the absolute value of the difference between the actual action of the patient and the standard action corresponding to the same joint part. As a specific example, for the patient's hip joint, the relevant data of the hip joint of the standard action in the rehabilitation training database is compared frame by frame, specifically: , Indicates the joint angle deviation value corresponding to the hip joint, is the joint angle in the actual motion data corresponding to the patient's hip joint, Represents the joint angle in the standard motion data corresponding to the hip joint.

[0038] Specifically, in the second aspect, the time domain features of the electromyographic signal are extracted and compared with the standard electromyographic signal to calculate the muscle activation difference. That is, based on the absolute value of the difference between the mean square error of the actual electromyographic signal data of the patient and the mean square error of the standard electromyographic signal data, the muscle activation difference value of the patient is obtained. As a specific example, the muscle activation difference value It can be expressed as: ,in represents the mean square error of the patient’s actual electromyographic signal data, Represents the mean square error of the standard electromyographic signal data.

[0039] Furthermore, the joint angle deviation value and the muscle activation difference value are weighted and summed to obtain the patient's comprehensive deviation score; and the actual deviation result includes two aspects of deviation calculation, namely the joint angle deviation value and the muscle activation difference value. Among them, the comprehensive deviation score reflects the patient's overall movement and muscle activation effect. The size of the weight can be set by the implementer according to the goal of rehabilitation. The larger the value of the comprehensive deviation score, the greater the deviation between the patient's current movement and the standard.

[0040] Then, the comprehensive deviation score is used to determine whether the patient's movement state meets the standard. That is, when the comprehensive deviation score is greater than or equal to the preset evaluation threshold, the patient's movement state does not meet the standard; when the comprehensive deviation score is less than the preset evaluation threshold, the patient's movement state meets the standard.

[0041] In other embodiments, a support vector machine (SVM) classification model can also be used to input the deviation score and electromyographic features into a classifier to identify whether the current action state meets the standard. The classification result and deviation data are output to be fed back to the closed-loop control module to guide the adjustment of the electrical stimulation parameters.

[0042] The control module is used to adjust the control parameters of the electrical stimulation according to the actual deviation results of the patient when the movement state of the patient does not meet the standard.

[0043] The control module receives the results of deviation analysis from the data collected by inertial motion capture and electromyography sensors, and dynamically adjusts the electrical stimulation parameters through closed-loop feedback to optimize the activation effect of the patient's muscles. In other words, the control module dynamically updates the electrical stimulation parameters based on electromyography signal feedback and motion status through an adaptive control algorithm.

[0044] In this embodiment, a closed-loop control algorithm (such as PID control, adaptive control, etc.) is used to adjust the start time, frequency and intensity of electrical stimulation in real time. If it is detected that the patient's movement deviates greatly from the standard movement, the system will appropriately increase the stimulation intensity or adjust the start time to make the movement closer to the standard trajectory.

[0045] Specifically, the PID controller is used to adjust the frequency, pulse width and amplitude of the electrical stimulation based on the actual deviation results of the patient to determine the control parameters. The output formula of the PID controller is: in, represents the output signal of the PID controller, Indicates the joint angle deviation value or muscle activation difference value, , and Respectively represent the proportional, integral and differential gain coefficients, and the optimal values ​​are obtained through empirical adjustment or online adaptive adjustment.

[0046] It should be noted that this embodiment dynamically adjusts the PID control parameters according to the fluctuation amplitude of the electromyographic feedback signal to meet the rehabilitation needs of different patients. , and The value of makes the electrical stimulation more in line with muscle needs.

[0047] Furthermore, in order to enhance the foresight and dynamic responsiveness of the system in this embodiment, model predictive control (MPC) is introduced. Specifically, based on the joint angles contained in the actual motion data currently collected by the patient and the electromyographic signals contained in the actual electromyographic data, the predicted motion data corresponding to the patient's movement state at several moments in the future are predicted. Using the optimization algorithm of model predictive control, constraints are set. The constraints may be that the comprehensive deviation score corresponding to muscle activation does not exceed the safety threshold, the joint angle contained in the predicted motion data reaches the target value, and so on. Furthermore, based on the model prediction results, model predictive control can optimize the frequency, pulse width and intensity of electrical stimulation in real time to ensure that the patient's movements gradually reach the standard while avoiding excessive muscle stimulation or fatigue.

[0048] The adjusted electrical stimulation parameters can be applied to the corresponding muscle groups, and personalized electrical stimulation can be performed for specific muscle groups in each step, such as activation of the hip flexors during the swing phase, stabilization of the quadriceps during the stance phase, etc. The system continuously fine-tunes the intensity and timing of electrical stimulation based on real-time feedback to achieve smooth and standardized walking, helping patients gradually restore normal gait.

[0049] Finally, it should be noted that in order to establish stable stimulation parameters, in this embodiment, the rehabilitation training system is optimized by accumulating and analyzing training data, optimizing parameters and performing stability tests, and the like.

[0050] Specifically, the adjusted electrical stimulation parameters are applied to the patient's target muscle groups (such as the adductor and abductor muscles of the thigh), and the spinal cord electrical stimulation device is used to guide the hip joint to complete the standard movement of adduction or abduction. The dynamic adjustment of the personalized electrical stimulation program ensures that the stimulation effect in each training can reach the best state.

[0051] During the rehabilitation training, the patient's movement trajectory and electrical stimulation parameters are continuously recorded to accumulate a large amount of training data. The control device analyzes this data, identifies effective stimulation parameter combinations, and summarizes the electrical stimulation intensity and frequency suitable for the patient through data analysis.

[0052] A machine learning model is used to gradually optimize the electrical stimulation parameters to ensure their stability and effectiveness in multiple training sessions. Through multiple rounds of training and feedback, the system optimizes the combination of electrical stimulation parameters without changing the patient's comfort, making the training effect more stable. When the system detects that the electrical stimulation parameters can stably guide the patient to complete the standard hip adduction and abduction movements, these parameters are set as the final stimulation plan for direct use in subsequent training to ensure continuous optimization of the patient's rehabilitation effect.

[0053] The user interaction adjustment module is used for patients and rehabilitation therapists to view the movement assessment results and electromyographic feedback data through the user interface, and is also used for patients and rehabilitation therapists to manually adjust the control parameters of electrical stimulation.

[0054] Patients and rehabilitation therapists can view the action assessment results and electromyographic feedback data in real time through the user interface. If the stimulation intensity is insufficient or too strong, the electrical stimulation parameters can be adjusted manually. The system will then update the parameter settings and enter the next motion capture and feedback process. The user interface can be viewed through a display device such as a tablet. Figure 3 It shows a schematic diagram of the user interface.

[0055] In summary, if Figure 2 It shows a flow chart of the closed-loop control of this embodiment. On the first hand, at the technical level, the present invention uses an integrated inertial motion capture device and a surface electromyography sensor to capture the patient's joint posture and muscle activation state in real time. The control module adopts an adaptive control algorithm and a closed-loop feedback mechanism to adjust the electrical stimulation parameters according to the patient's real-time data, thereby ensuring that the electrical stimulation is synchronized with the patient's motion state, thereby improving the accuracy of the electrical stimulation and the rehabilitation effect. By analyzing the patient's real-time motion and electromyography feedback data, a personalized electrical stimulation program can be provided to meet the needs of patients at different rehabilitation stages. Compared with traditional fixed-parameter electrical stimulation, the personalized adjustment mode of the present invention can greatly improve the pertinence and effect of training, shorten the rehabilitation cycle, and enhance the patient's recovery ability. By fusing inertial and electromyography data, the accuracy of motion capture and the reliability of data are improved, and the problem of single sensor data being susceptible to interference is avoided. Through advanced data fusion algorithms, the system can more stably obtain high-precision patient motion data, providing solid data support for closed-loop control.

[0056] On the second aspect, at the application level, the present invention can significantly improve the rehabilitation effect of patients with spinal cord injury by adjusting electrical stimulation parameters in real time and personalized closed-loop control. Real-time feedback makes the patient's training effect more stable and the rehabilitation progress more effective. At the same time, personalized training programs can reduce patients' discomfort with rehabilitation training and improve rehabilitation efficiency. Integrating inertial motion capture and electromyography sensors on training clothes realizes the integration of data acquisition equipment and the convenience of wearing. Patients can complete rehabilitation training without wearing multiple separate devices, which not only increases the comfort of training, but also improves the flexibility and practicality of the rehabilitation process. The integrated design and dynamic electrical stimulation control method are not only suitable for rehabilitation training of spinal cord injury, but can also be widely used in other rehabilitation scenarios that require electrical stimulation assistance, such as rehabilitation of diseases such as stroke and muscular atrophy. The modular design of the system makes it have good adaptability and scalability, and can meet a variety of rehabilitation needs.

[0057] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. An integrated rehabilitation training system based on motion capture and electromyographic feedback, characterized in that: The system includes: The motion data acquisition module is used to collect the actual motion data and actual electromyographic signal data of the patient in the motion state, and obtain the standard motion data and standard electromyographic signal data of the standard motion in the rehabilitation training database; A motion evaluation and analysis module, used to identify the motion state according to the actual motion data and actual electromyographic signal data of the patient, and determine the actual motion of the patient; According to the difference between the actual motion data of the actual action of the patient and the standard motion data of the standard action, and the difference between the actual electromyographic signal data of the actual action and the standard electromyographic signal data of the standard action, a comprehensive deviation score and an actual deviation result of the patient are obtained; and whether the movement state of the patient meets the standard is judged according to the comprehensive deviation score; The control module is used to adjust the control parameters of the electrical stimulation according to the actual deviation results of the patient when the movement state of the patient does not meet the standard.

2. The integrated rehabilitation training system based on motion capture and electromyographic feedback according to claim 1, characterized in that: The collecting of actual motion data of the patient in motion state specifically includes: The inertial detection unit in the training suit collects the three-axis acceleration and three-axis gyroscope data of each joint of the patient; the sampling frequency is at least 100Hz; Based on the three-axis acceleration and three-axis gyroscope data of various parts of the patient, the posture angle of the corresponding joint is calculated to obtain the joint angle and angular velocity of various parts of the patient; the patient's actual motion data includes the joint angle and angular velocity of various parts of the patient.

3. The integrated rehabilitation training system based on motion capture and electromyographic feedback according to claim 2, characterized in that: The method for obtaining the patient's comprehensive deviation score and actual deviation result specifically includes: Based on the absolute value of the difference between the actual action of the patient and the standard action of the patient at the same joint position, a joint angle deviation value of the patient is obtained; Obtaining a muscle activation difference value of the patient based on an absolute value of a difference between a mean square error of the actual electromyographic signal data of the patient and a mean square error of the standard electromyographic signal data; The joint angle deviation value and the muscle activation difference value are weightedly summed to obtain a comprehensive deviation score for the patient; the actual deviation result includes the joint angle deviation value and the muscle activation difference value.

4. The integrated rehabilitation training system based on motion capture and electromyographic feedback according to claim 3, characterized in that: The step of adjusting the control parameters of the electrical stimulation according to the actual deviation result of the patient specifically includes: The PID controller is used to adjust the frequency, pulse width and amplitude of the electrical stimulation based on the actual deviation results of the patient to determine the control parameters.

5. The integrated rehabilitation training system based on motion capture and electromyographic feedback according to claim 4, characterized in that: The output formula of the PID controller is: in, represents the output signal of the PID controller, Indicates the joint angle deviation value or muscle activation difference value, , and Represent the proportional, integral and differential gain coefficients respectively.

6. The integrated rehabilitation training system based on motion capture and electromyographic feedback according to claim 1, characterized in that: The step of judging whether the patient's movement state meets the standard according to the comprehensive deviation score specifically includes: When the comprehensive deviation score is greater than or equal to the preset evaluation threshold, the patient's motor status does not meet the standard; When the comprehensive deviation score is less than the preset evaluation threshold, the patient's movement status meets the standard.

7. The integrated rehabilitation training system based on motion capture and electromyographic feedback according to claim 1, characterized in that: The system also includes a user interaction adjustment module, which is used for patients and rehabilitation therapists to view movement evaluation results and electromyographic feedback data through a user interface, and is also used for patients and rehabilitation therapists to manually adjust control parameters of electrical stimulation.

Images