A motion rehabilitation coordination evaluation system based on electromyographic similarity

Abstract

The application discloses a motion rehabilitation coordination evaluation system based on electromyography similarity, which comprises a collection module, a processing module, a training module and a display module. The display module displays a virtual upper limb and a target state gap to drive the real upper limb movement; the processing module judges the movement ability based on the gap change and controls the auxiliary driving. The processing module evaluates the coordination based on the similarity of the bilateral upper limb electromyography signals; and based on the difference value of the first and second electromyography signals, the size of the surplus value is adjusted by adjusting the preset proportion, and then the total value of the driving torque applied to one side of the upper limb is dynamically compensated. The application realizes precise control of the driving torque, avoids the uncoordinated limb recovery caused by unsuitable training intensity, and improves the rehabilitation effect of stroke patients.

Description

[0001] The original basis for this divisional application is patent application No. 202310978929.0, filed on August 4, 2023, entitled "An Upper Limb Rehabilitation Assessment System for Stroke Patients". Technical Field

[0002] This invention relates to the field of stroke rehabilitation technology, and in particular to a motor rehabilitation coordination assessment system based on electromyographic similarity. Background Technology

[0003] Among stroke patients, there are hemiplegic patients with partial paralysis of either side of their limbs, and systems to assist these patients in rehabilitation training are becoming increasingly widespread. Existing upper limb paralysis rehabilitation training includes at least assistive motor rehabilitation training using upper limb rehabilitation robots and exoskeletons. Rehabilitation robots can provide high-intensity, repetitive, task-specific, and interactive treatment for the paralyzed upper limbs, while also providing objective evaluation methods for motor function recovery, measuring kinematic and biomechanical changes. However, upper limb rehabilitation robots cannot improve upper limb muscle strength. Motor function assessment is the basis for doctors to develop rehabilitation training plans for stroke hemiplegic patients. It can also serve as a means of evaluating the effectiveness of various treatment options and is crucial in the fields of motor rehabilitation and neuroscience. Traditional motor rehabilitation assessments are mainly based on scales and are conducted by professional therapists. Rating scales have been developed relatively early and their feasibility has been clinically validated. However, they have three drawbacks: First, the assessment results are significantly influenced by the therapist's subjectivity. Different therapists assess based on experience and use different standards, potentially leading to different results for the same patient. Second, rating scales assign scores based on the patient's different movements, resulting in small scoring gradients and inaccurate assessments that don't reflect the patient's true condition. Furthermore, therapists may struggle to visually capture subtle movements, leading to incomplete assessments. Third, current rating scales require a therapist to guide the patient through a series of movements, necessitating at least one therapist per patient, resulting in low efficiency. Finally, the entire assessment process is time-consuming and labor-intensive.

[0004] Existing technologies, such as the patent document with publication number CN110755085B, propose a method and device for assessing motor function based on joint range of motion and motor coordination. The method includes: acquiring limb movement signals of a target object; extracting joint range of motion data and motor coordination data from the limb movement signals; and assessing the motor function of the limb, wherein the joint range of motion data includes active movement data and passive movement data. Existing rehabilitation training assessment systems typically use limb response values ​​from healthy individuals as training standards. However, for stroke patients, their limb training response cannot reach normal standards in a short period. In particular, most stroke patients have hemiplegia, leading to significant differences in training response between the left and right limbs. If a universal standard is directly applied for training and assessment, providing the same training intensity to both limbs, the less responsive limb will bear excessive training pressure, not only failing to achieve rehabilitation goals but also potentially causing greater damage. Furthermore, such indiscriminate training slows down the patient's rehabilitation process, hindering their normal life.

[0005] Furthermore, on the one hand, there are differences in understanding among those skilled in the art; on the other hand, the inventors studied a large number of documents and patents when making this invention, but due to space limitations, not all details and contents were listed in detail. However, this does not mean that the present invention does not possess the features of these prior art. On the contrary, the present invention already possesses all the features of the prior art, and the applicant reserves the right to add relevant prior art to the background art. Summary of the Invention

[0006] To address the shortcomings of existing technical solutions, this application proposes an upper limb rehabilitation assessment system for stroke patients, comprising at least: a data acquisition module for acquiring upper limb information from the stroke patient; a processing module for processing the upper limb information acquired by the acquisition module; a training module for assisting the stroke patient in upper limb movement; and a display module for displaying the upper limb movement status of the stroke patient. The display module shows the difference between the virtual upper limb and the target state to the stroke patient, enabling the stroke patient to drive the real upper limb to move. The processing module determines the motor ability of the real upper limb based on the change in the difference between the virtual upper limb and the target state, and controls the training module to perform assisted movement according to the motor ability. The processing module adjusts the driving torque applied to both upper limbs by the training module based on the difference in electromyographic signals between the two upper limbs acquired by the acquisition module. The processing module assesses the motor ability and coordination of both upper limbs and provides targeted rehabilitation assistance.

[0007] In the above scheme, the virtual upper limb refers to the upper limb model constructed in the display module, the real upper limb refers to the patient's upper limb in the actual environment, the target state refers to the upper limb pattern that is randomly generated in the display module to allow the patient to move the upper limb to the target position and the target posture, and the gap between the virtual upper limb and the target state refers to the gap between the current position of the virtual upper limb and the target position, as well as the gap between the current posture of the virtual upper limb and the target posture.

[0008] This application focuses on the differences in limb movement between the left and right sides of the patient, and provides different and targeted training parameters for each side. While improving the training effect and efficiency, it also avoids the situation where the muscles of the left and right limbs become uncoordinated during the rehabilitation process due to the application of inappropriate training intensity. It solves the problem that the lack of coordination between the two sides of the limbs can hinder the normal life of patients, and has a certain positive impact on the quality of life of stroke patients.

[0009] Preferably, the processing module controls the driving torque of the training module to increase gradually and slowly, and the training module stops increasing the torque when the real upper limb shows a tendency to move, and provides auxiliary driving force to the real upper limb with a surplus value of the current driving torque plus a preset proportion of the current driving torque. In this mode, the actual driving force of the real upper limb movement is still mainly driven by the stroke patient's own initiative, with the training module driving force as a supplement. The increased surplus value can avoid muscle overuse and protect the muscle health of the stroke patient.

[0010] Preferably, the processing module adjusts the surplus value based on the difference between the first and second electromyographic (EMG) signals of both upper limbs acquired by the acquisition module, using a preset ratio, thereby adjusting the total value of the driving torque applied to one upper limb. The first and second EMG signals refer to the EMG signals acquired by the EMG acquisition unit of the acquisition module during the movement of one of the patient's upper limbs. Specifically, the first EMG signal refers to the EMG signal generated by the movement of the left upper limb, and the second EMG signal refers to the EMG signal generated by the right upper limb. The total value of the driving torque is equal to the driving torque at the point of movement, plus a surplus value calculated according to a preset ratio (e.g., one-tenth) of the driving torque.

[0011] Assuming the intensity of the first electromyographic signal in the left upper limb is greater than the intensity of the second electromyographic signal in the right upper limb, indicating that the recovery of the left upper limb is better than that of the right upper limb, the driving torque applied to the right upper limb should be smaller than that applied to the left upper limb when the training module applies driving torque. The design logic of this approach is that by applying a smaller driving torque to the right upper limb, which has a poorer recovery, the right upper limb receives more exercise. Furthermore, because a smaller driving torque is applied, the stroke patient's attention will be more focused on driving the right upper limb during the movement, thereby gradually bringing the recovery effect of the right upper limb to the same level as that of the left upper limb. This achieves the effect of maintaining coordination between the two upper limbs during the rehabilitation training process.

[0012] Preferably, the processing module is at least capable of constructing a virtual upper limb in the virtual environment of the display module based on the electromyographic signals and motion information of the real upper limb acquired by the acquisition module, wherein the acquisition module includes at least an electromyographic acquisition unit for acquiring electromyographic signals and a motion capture unit for capturing movements.

[0013] Preferably, the step of the processing module mapping a real upper limb to a virtual upper limb in the virtual environment includes: To build a virtual upper limb model in a virtual environment; Collect electromyographic signals and posture information of the real upper limb; Feature extraction is performed on electromyographic signals, motion information features, and posture information of the real upper limb, and then the features are classified in a hierarchical manner according to their strength. The electromyographic features and the angular features of the motion are fused and then input into the classifier for pattern recognition. It dynamically recognizes the movements and motions of the real upper limbs and synchronously adjusts the movements and motions of the virtual upper limbs.

[0014] Preferably, the display module is configured in a way that can be observed by stroke patients, so that stroke patients can continuously imagine the virtual upper limb moving to the target state while observing the image of the virtual upper limb and the target state in the display module.

[0015] Preferably, the processing module can analyze the time and movement state of the virtual upper limb in the display module as it moves to the target state, and judge the movement ability of the real upper limb in the actual environment based on the time and movement state.

[0016] Preferably, the processing module can perform inverse kinematics analysis on the driving angle and driving torque required for the training module to perform assisted movement, wherein the driving torque calculated by the processing module is the maximum driving torque that the training module can provide under the current assisted movement.

[0017] Preferably, the steps of the processing module in calculating the driving angle and driving torque include: The current position of the virtual upper limb end and the target position of the target state are obtained, and the driving angles of each joint are calculated from the current position and the target position based on the inverse kinematics algorithm. The arm support torque required for each joint is dynamically adjusted based on iterative learning. The arm support torque required for each joint is calculated from the target driving angle and the actual joint angle, target joint angular velocity, and actual joint angular velocity. Inverse dynamics control is based on feedback linearization. The output torque of each joint is calculated from the target motion parameters and actual motion parameters of each joint, and the driving force of each joint is calculated from the output torque of each joint. The driving torque is applied to the joint to be trained according to the driving amount of each joint in order to carry out rehabilitation training.

[0018] Preferably, the electromyography (EMG) acquisition unit of the acquisition module and the training module are jointly configured on the bilateral upper limbs of the stroke patient, and the motion capture unit of the acquisition module and the display module are jointly configured.

[0019] According to a preferred embodiment, this application proposes an upper limb rehabilitation assessment system for stroke patients. This system can be applied to assess the rehabilitation effects of stroke patients in their homes. In existing stroke rehabilitation assessment technologies, patients typically need to visit a professional medical staff for various instrument and movement analysis tests to understand their rehabilitation status. This approach places a significant time and financial burden on ordinary families. Therefore, this application proposes a portable, home-use upper limb rehabilitation assessment system for stroke patients. This system displays various upper limb movements via a display module. Patients can then imagine performing the same movements with their own upper limbs. By analyzing and comparing the electromyographic signals generated by the upper limb movements, the system can assess the rehabilitation effect and the coordination between upper limbs during the rehabilitation process. By placing this system in the patient's home, stroke patients can use it at any time to assess their rehabilitation status and receive targeted rehabilitation training. Attached Figure Description

[0020] Figure 1 This is a simplified relational structure diagram of the upper limb rehabilitation assessment system for stroke patients of the present invention; Figure 2 This is a simplified structural diagram of the display module of the upper limb rehabilitation assessment system for stroke patients of the present invention; Figure 3 This is a simplified structural diagram of the training module of the upper limb rehabilitation assessment system for stroke patients of the present invention.

[0021] List of reference numerals 100: Acquisition module; 200: Processing module; 300: Training module; 400: Display module; 110: Electromyography acquisition unit; 120: Motion capture unit. Detailed Implementation

[0022] The following is in conjunction with the appendix Figure 1-3 The present invention will be described in detail below.

[0023] Figure 1 A simplified relational structure diagram of an upper limb rehabilitation assessment system for stroke patients according to this application is shown. The system includes at least a training module 300 for applying assisted movement, an acquisition module 100 for acquiring physiological information and movement status of both upper limbs, a display module 400 for displaying the target state, and a processing module 200 for analysis and processing. The acquisition module 100 is capable of acquiring the movement status of both upper limbs and electromyographic signals after movement. The processing module 200 virtually displays the patient's upper limbs on the display module 400 based on the movement status for observation by the stroke patient. Furthermore, the display module 400 internally displays the target state. With the assistance of the training module 300, the stroke patient needs to move both upper limbs to the target state according to the observed target state.

[0024] The technique of using a display module 400 to allow stroke patients to observe the movement status of both upper limbs is an extension of motor imagery therapy and mirror therapy. Motor imagery therapy refers to repetitive motor imagery performed to improve motor function without any actual motor output. Based on motor memory, it activates specific areas of the brain for a particular activity. Compared to actual movement, motor imagery activates more of the anterior frontal lobe and posterior parietal lobe. Mirror therapy involves placing a mirror in a position convenient for the patient to observe. The non-paralyzed upper limb is placed in front of the mirror, and the affected upper limb is placed behind it. The non-paralyzed upper limb performs flexion and extension movements of the wrist and fingers, while the patient simultaneously looks in the mirror to observe the mirror image of the non-paralyzed upper limb. The paralyzed upper limb performs the same movements. By regulating cortical-muscle excitability, mirror therapy can directly promote the recovery of motor function. Mirror therapy is a type of motor imagery therapy based on repetitive motor function imagery.

[0025] In this application, according to Figure 2The simplified structural diagram of the display module 400 of the stroke patient upper limb rehabilitation assessment system of the present invention is shown. The movement of the non-paralyzed upper limb used for comparison in mirror therapy is set as the target state. That is, the display module 400 is equivalent to the mirror in mirror therapy. The target state in the display module 400 is the non-paralyzed upper limb in mirror therapy. By observing the simulated upper limb movement state in the display module 400, stroke patients can imagine in their brain how their actual bilateral upper limbs should move so that the simulated virtual upper limbs in the display module 400 can move from the current position to the target state position. This promotes the excitability and correlation between the cerebral cortex and the upper limb muscles, thereby assisting stroke patients with upper limb paralysis in rehabilitation training.

[0026] Preferably, the acquisition module 100 can acquire upper limb movement state information in the actual environment and process and analyze it through the processing module 200, and then display it in the form of a virtual upper limb in the display module 400. The virtual upper limb in the display module 400 can synchronously adjust the state of the virtual upper limb based on the real limb movement state information acquired in real time by the acquisition module 100.

[0027] Preferably, the processing module 200 can accurately analyze the movement trajectory of the virtual upper limb in the display module 400. Specifically, the processing module 200 can determine what kind of movement trajectory the virtual upper limb needs to make to move from its current position to the target state position, and then determine the auxiliary movement parameters of the training module 300 to assist the upper limb movement in the actual environment, so that the upper limb can complete the target state.

[0028] Preferably, the target state in the display module 400 can be static or dynamic. A static target state refers to a specific posture, such as an extended arm, a bent arm, or a clenched fist. A dynamic target state refers to the movements that the arm can perform, such as: Finger extension: accomplished by the coordinated action of the palmar extensor muscles, finger extensor muscles, and thumb abductor muscles; Finger flexion: accomplished by the coordinated action of the flexor muscles of the fingers, thumb, and wrist. Clenching a fist: is accomplished by the action of muscles such as the palmar flexor muscles, finger extensor muscles, and thumb abductor and extensor muscles. Spreading fingers: This is accomplished by the action of muscles such as the palmar extensor muscles, finger extensor muscles, and thumb abductor and extensor muscles. Rotating the palm: This is accomplished by the coordinated action of the forearm rotator muscles, wrist flexors, and wrist extensors. Raising the wrist: This is accomplished by the action of muscles such as the wrist flexor, radial extensor, and ulnar extensor. The electromyographic signals of the real upper limbs in the above-mentioned actions and postures as target states are acquired by electromyographic sensors located in the corresponding muscle areas.

[0029] Preferably, the acquisition module 100 includes an electromyography (EMG) acquisition unit 110 and a motion capture unit 120. The EMG acquisition unit 110 is used to acquire EMG signals and other information under different motion states and send them to the processing module 200. The processing module 200 constructs the muscle changes of the virtual upper limb within the display module 400 based on the EMG signals and other information acquired by the EMG acquisition unit 110. The motion capture unit 120 is used to acquire the joint coordinates and spatial posture information of the upper limb and send them to the processing module 200. The processing module 200 adjusts the coordinates and posture of the virtual upper limb in the display module 400 based on the changes in joint coordinates and spatial posture information.

[0030] Preferably, the electromyography (EMG) acquisition unit 110 employs a surface electromyography sensor, EMG electrodes, and / or an EMG armband. The EMG acquisition unit 110 can be used to acquire EMG signals and other information when both upper limbs are in different movement states in a real environment. Furthermore, the portable EMG armband can also acquire displacement signals of the upper limbs of stroke patients. Specifically, the EMG armband can be worn on at least the forearm and upper arm of stroke patients respectively, and can acquire bioelectrical signals generated by the muscles of the forearm and upper arm, as well as motion acceleration and displacement signals generated by movement.

[0031] Preferably, the display module 400 essentially provides a virtual environment, wherein the virtual upper limb and the target state position are both located within the virtual environment of the display module 400. The virtual environment of the display module 400 is built using existing technologies, such as Unity3D or other software technologies capable of 3D modeling. This technology has been widely used in the field of virtual modeling, so it will not be described in detail here.

[0032] Preferably, the processing module 200 maps the upper limb information acquired by the acquisition module 100 onto the virtual upper limb of the display module 400, specifically including the mapping of the shoulder, elbow and finger joints in the upper limb.

[0033] Preferably, in this application, the processing module 200 constructs a virtual upper limb capable of synchronizing with the upper limb movement and posture information acquired by the electromyography acquisition unit 110 and the motion capture unit 120 of the acquisition module 100 in the virtual environment of the display module 400, specifically including the following steps: S110: Construct a virtual upper limb model in a virtual environment; map the real upper limb in the actual environment to the virtual upper limb in the virtual environment of the display module 400; map the motion information of the real upper limb collected by the electromyography acquisition unit 110 and the motion capture unit 120 to the virtual upper limb in the virtual environment. S120: Acquires electromyographic signals and posture information of the real upper limb; Acquires electromyographic signals and posture information of the upper limb through the electromyographic acquisition unit 110 and the motion capture unit 120; S130: Extract features from electromyography signals, motion information features, and posture information of the real upper limb, and classify them in a hierarchical manner according to the strength of the features. S140: After fusing electromyographic features and angular features of motion, the data is input into the classifier for pattern recognition; S150: Dynamically recognizes the movements and motions of the real upper limbs and synchronously adjusts the movements and motions of the virtual upper limbs; establishes communication between the virtual environment and mathematical software such as MATLAB, recognizes the movements and motions of the real upper limbs online, and synchronously adjusts the movements and motions of the virtual upper limbs in the display module 400.

[0034] Preferably, in step S110, the mapping of the virtual upper limb includes mapping the motion information of the hand, forearm, and upper arm of the real upper limb captured by the motion capture unit 120 to the hand, forearm, and upper arm of the virtual upper limb in the virtual environment, assigning the preset component of the myoelectric arm ring to the upper arm of the virtual upper limb in the virtual environment, synchronizing the motion information of the upper arm, assigning the posture and position information of each joint of the fingers and wrist joint captured by the hand image motion capture unit 120 to the fingers and forearm of the virtual upper limb, synchronizing the motion information of the fingers and forearm, calibrating the initial orientation of the upper arm and forearm in the virtual environment, and adjusting the virtual upper limb to synchronize with the real upper limb.

[0035] Preferably, in step S120, the acquisition module 100 includes electromyographic signals acquired by the electromyographic acquisition unit 110 or electromyographic electrodes, and the motion capture unit 120 of the acquisition module 100 is set at the display module 400 to acquire gesture information and forearm and upper arm posture information through the motion capture unit 120. The portable electromyographic armband communicates with the processing module 200 through a Bluetooth receiver. The processing module 200 sets the electromyographic acquisition method in a virtual environment to acquire forearm electromyographic information when performing different gestures, and establishes a gesture library of multiple gestures. The acquisition method includes a process of preparing a gesture countdown - executing the gesture - repeating the action. The posture and position information of the fingers and forearm captured by the hand image motion capture unit 120 is used to calculate angle information such as finger bending angle and elbow joint range of motion. The upper arm movement information acquired by the portable electromyographic armband is used to calculate shoulder joint range of motion.

[0036] Preferably, in step S130, feature extraction includes extracting time-domain features of electromyography (EMG) signals, including mean absolute value, number of zero-crossing points, and coefficients of the autoregressive model; extracting gesture information features such as joint range of motion; and extracting upper limb posture information such as position information. A spatial motion trajectory is set, and the degree of trajectory deviation is calculated. The EMG features and posture information of a normal person are used as templates, and a hierarchical classification is performed according to feature strength. The user's EMG features and posture information are then assigned corresponding scores according to the corresponding hierarchical categories.

[0037] Preferably, in step S140, pattern recognition includes fusing electromyographic features and angular features of gestures, then inputting them into a classifier such as a BP neural network for pattern classification, saving the parameters and thresholds of the virtual upper limb, and using position information and other posture information as conditions to determine the direction of movement.

[0038] Preferably, in step S150, online identification specifically involves: calling MATLAB functions in the processing module 200 or establishing communication between the virtual environment in the display module 400 and MATLAB; then identifying and classifying the stroke patient's movements in the processing module 200; and driving the virtual upper limb to perform synchronous movements in the virtual environment based on the classification.

[0039] Preferably, the processing module 200 can analyze the time it takes for the virtual upper limb in the display module 400 to move to the target state, judge the actual upper limb's movement ability in the actual environment based on the time and movement state, and provide a certain amount of assistance and support through the training module 300 when the movement ability is insufficient, thereby helping stroke patients complete the movement from the initial state to the target state.

[0040] Example 2 This embodiment is an improvement and supplement to Embodiment 1, and repeated content will not be repeated.

[0041] like Figure 1 As shown, an upper limb rehabilitation assessment system for stroke patients includes at least a data acquisition module 100, a processing module 200, a training module 300, and a display module 400. The display module 400 is configured in a manner observable by the stroke patient and displays the target state of upper limb movement and a virtual upper limb. The virtual upper limb is mapped by the processing module 200 onto the virtual environment of the display module 400 based on the upper limb information acquired by the data acquisition module 100. During training, the processing module 200 can judge the actual upper limb's movement ability in the real environment based on the movement time and movement state of the virtual upper limb in the virtual environment of the display module 400. When the processing module 200 determines that the actual upper limb's movement ability is insufficient to complete the movement from the initial state to the target state, the processing module 200 sends a control signal to the training module 300, which then assists the actual upper limb in completing the movement in the real environment.

[0042] Preferably, the processing module 200 mainly makes judgments based on the difference between the virtual upper limb and the target state in the virtual environment of the display module 400 and the time. For example, if the virtual upper limb in the display module 400 does not move to the preset position of the target state within a certain preset time period, or if the gesture is not accurate after moving to the preset position, etc. Specifically, the processing module 200 can rate the motor ability of the virtual upper limb (the virtual upper limb is synchronized with the real upper limb, which essentially refers to the real upper limb) based on the difference between the virtual upper limb and the target state. More specifically, during the rating process, the processing module 200 can also refer to the electromyography (EMG) signal of the real upper limb acquired by the EMG acquisition unit 110 of the acquisition module 100. By analyzing the EMG signal, the motor ability of the real upper limb can be judged to a certain extent.

[0043] Preferably, when the processing module 200 considers the motor ability of the real upper limb (virtual upper limb) to be insufficient, the processing module 200 analyzes the driving angle and driving torque required by the training module 300 based on inverse kinematics.

[0044] Preferably, for the movement state of the upper limb, the various movements of the entire upper limb can be decomposed into the rotational movements of each joint. Specifically, no matter what state or form of normal movement the upper limb is in, the length and shape of its bones will not change. The flexibility of the upper limb mainly depends on the multi-angle flexible rotation of the joints connecting each bone. Therefore, when applied to the exoskeleton of the training module 300, it is only necessary to control the rotation angle and rotation direction of the joint connection between each bone to achieve upper limb assisted movement.

[0045] Preferably, the steps by which the processing module 200 calculates the driving angle and driving torque of the training module 300 based on the distribution of the virtual upper limb and target state in the virtual environment of the display module 400 mainly include: S210: Obtain the current position of the virtual upper limb end and the target position of the target state corresponding to the rehabilitation training task, and calculate the driving angle of each joint based on the current position and the target position using the inverse kinematics algorithm; S220: Based on iterative learning, the arm support torque of each joint of the online adjustment training module 300 is calculated from the target driving angle and the actual joint angle, target joint angular velocity, and actual joint angular velocity to obtain the arm support torque required for each joint. S230: Inverse dynamics control is performed based on feedback linearization. The output torque of each joint is calculated from the target motion parameters and actual motion parameters of each joint, and the driving amount of each joint is calculated based on the output torque of each joint. S240: Control the application of driving torque to the joint to be trained according to the driving amount of each joint in order to carry out rehabilitation training.

[0046] Preferably, the training module 300 is based on the driving angle and driving torque calculated by the processing module 200, wherein the calculated driving torque is the maximum driving torque that the training module 300 can provide under this assisted movement, thereby assisting the upper limbs of stroke patients to move.

[0047] Preferably, the driving torque of the training module 300 is configured to increase gradually and slowly, and to stop increasing when the real upper limb shows a tendency to move, and to provide auxiliary driving to the real upper limb with a surplus value of the current driving torque plus a preset ratio of the current driving torque. For stroke patients who are not completely paralyzed, a better rehabilitation training method is to allow them to drive their upper limbs to move on their own. The technical solution of this application is designed based on this method. That is, the driving torque applied by the training module 300 increases from zero. When the real upper limb is moving, the driving torque applied by the training module 300 remains unchanged. Only when the real upper limb does not move for a period of time does the driving torque of the training module 300 gradually increase, and the rate of increase is relatively slow. If the real upper limb shows a tendency to move during the increase, the driving torque of the training module 300 stops increasing and is increased by a preset ratio (e.g., one-tenth) of the driving torque when the tendency to move appears. That is, in this way, the actual driving force of the real upper limb movement is still mainly driven by the stroke patient's own driving force, with the driving force of the training module 300 as a supplement. The increased surplus value can avoid muscle overuse and protect the muscle health of the stroke patient.

[0048] Compared to existing driving schemes that only provide driving force, the training module 300 of this application focuses on assisting stroke patients in moving their bilateral upper limbs. During the movement, the stroke patient watches the movement state and target state of the virtual upper limb in the display module 400, and continuously imagines the way to move the virtual upper limb to the target state. This stimulates the brain signals to continuously send driving signals to the neurons connected to the muscles, so that the muscles contract as much as possible. When the muscles contract to the maximum extent that they can contract, the training module 300 is used to assist the upper limbs in movement, so that the brain can remember the feedback signals of muscle movement, strengthen the control connection between the brain and the upper limb muscles in rehabilitation training, and thus improve the efficiency of rehabilitation training.

[0049] Example 3 This embodiment is an improvement and supplement to Embodiments 1 and 2, and repeated content will not be repeated.

[0050] In reality, most daily activities require the participation of both upper limbs. Unilateral upper limb paralysis severely impairs patients' ability to perform functional activities requiring the participation of both upper limbs. Unilateral upper limb activities and bilateral upper limb activities have different neural control mechanisms. Only through simultaneous training of both upper limbs can patients achieve the ability to engage both limbs simultaneously. For example, one upper limb can be used to fix an object, while the other can be used to manipulate objects, such as twisting a bottle cap or chopping vegetables. Simultaneous training of both upper limbs can improve the function of the paralyzed upper limb, especially showing more beneficial progress in proximal upper limb function. This may be because the trunk and proximal limb muscles are innervated bilaterally, or it may be that bilateral symmetrical training generates greater electromyographic activity in the trunk muscles, thus promoting trunk stability, which is crucial for proximal limb control. During training, it is important to select different upper limb tasks based on the patient's specific condition.

[0051] Therefore, in actual rehabilitation treatment, the differences in recovery between the two upper limbs need to be considered. If assisted training is performed with the same driving torque on both sides, the degree of recovery of the two upper limbs may differ. Whether a difference exists can be determined by the electromyographic signals acquired by the acquisition module 100. If the same driving torque is still used for assisted rehabilitation training of the two upper limbs when a difference exists, the difference in the degree of recovery between the two upper limbs may become larger and larger, resulting in incoordination between the two upper limbs.

[0052] Therefore, this application proposes the following: Figure 1 The system shown is an upper limb rehabilitation assessment system for stroke patients. The system includes at least a data acquisition module 100, a processing module 200, a training module 300, and a display module 400. The processing module 200 can calculate the driving angle and driving torque that the training module 300 needs to apply to the real upper limb for assisting movement based on the distance between the target state in the display module 400 and the virtual upper limb. In addition, the processing module 200 can also adjust the driving torque applied to the bilateral upper limbs based on the difference between the first electromyographic signal and the second electromyographic signal of the bilateral upper limbs acquired by the electromyographic acquisition unit 110 of the data acquisition module 100.

[0053] Specifically, assuming that the electromyography (EMG) signal of the left upper limb of the stroke patient acquired by the EMG acquisition unit 110 of the acquisition module 100 is recorded as the first EMG signal, and the EMG signal of the right upper limb of the stroke patient acquired by the EMG acquisition unit 110 is recorded as the second EMG signal, the processing module 200 analyzes the first EMG signal and the second EMG signal, calculates the difference value between the first EMG signal and the second EMG signal, and adjusts the driving torque applied to both upper limbs by the training module 300 according to the difference value.

[0054] Preferably, the processing module 200's analysis of electromyographic signals mainly includes raw surface electromyographic signal analysis and processed data analysis. Data analysis focuses primarily on time and frequency domain analysis. The purpose of signal analysis is to study the correlation between the time-frequency characteristics of surface electromyographic signals and muscle structure, activity state, and functional state, to explore the possible causes of surface signal changes, and to effectively utilize changes in electromyographic signals to reflect muscle activity and function. The raw surface electromyographic signal, as the most direct representation of the occurrence and resting state of electromyographic activity, can be analyzed for its initial relationship, i.e., the density and height of the raw electromyographic signal during muscle activity, without considering amplitude. This can reflect the amplitude and force of muscle contraction to a certain extent. Higher density and height indicate a stronger surface electromyographic signal and thus a stronger contraction. Processed data analysis involves using the built-in signal processing system in the software to rectify, smooth, and perform MVC normalization on the directly recorded raw surface electromyographic signal, followed by further calculation and analysis. For example, the analysis software ErgoLAB provides methods for analyzing raw surface electromyography signals as well as processed data such as time-domain analysis, frequency-domain analysis, and segmented analysis.

[0055] Preferably, the processing module 200 is at least capable of analyzing the electromyographic signals acquired by the acquisition module 100 to determine the muscle contraction state, and determining whether the muscle movement is normal based on a preset muscle contraction threshold. Specifically, the processing module 200 mainly monitors abnormal muscle contraction states, and determines whether the muscle is in an abnormal movement state by calculating the difference between the muscle contraction value obtained from the real-time acquired electromyographic signals and the preset abnormal muscle contraction threshold. It also monitors abnormal muscle contraction patterns that occur during rehabilitation training. These abnormal muscle contraction patterns include muscle overexcitation, muscle overinhibition, abnormal muscle synergy, abnormal co-contraction of agonist and antagonist muscles, and muscle fatigue.

[0056] Preferably, the processing module 200 can also determine the connection response between different muscles and EEG signals during movement based on the magnitude of electromyographic signals of different muscles in the same upper limb acquired by multiple electromyographic acquisition units 100 of the acquisition module 100. Specifically, the upper limb, from the shoulder to the fingers, includes the deltoid, biceps brachii, triceps brachii, brachialis, pronator teres, brachioradialis, extensor carpi radialis longus, extensor carpi radialis brevis, flexor carpi radialis, palmaris longus, flexor digitorum superficialis, and extensor carpi ulnaris. During upper limb movement, the contraction and extension of muscles in each part are not consistent. The processing module 200 analyzes the current movement of the upper limb based on the motion capture unit 120 to determine the appropriate electromyographic signal range for the muscles of the upper limb, and then applies this information to the electromyographic acquisition units 110 of the acquisition module. The system collects electromyographic signals from different parts of the body to determine the activity of each muscle during the rehabilitation process. For example, during arm flexion, the biceps, triceps, and brachialis should be relatively active. If the electromyographic signal of the triceps is detected to be too weak during this process, the processing module 200 considers that the triceps has a greater connection with the brain and gives the triceps a weak activity rating. In the subsequent rehabilitation process, the system will try to focus on adopting exercise postures that can train the triceps to comprehensively improve the rehabilitation effect.

[0057] Preferably, the processing module 200 analyzes the difference between the first electromyographic signal and the second electromyographic signal based on the above analysis method. Specifically, the processing module 200 evaluates the coordination of the bilateral upper limbs of the stroke patient based on the similarity between the first electromyographic signal and the second electromyographic signal, wherein the first electromyographic signal and the second electromyographic signal are electromyographic signals collected synchronously when the bilateral upper limbs perform the same action.

[0058] Preferably, the processing module 200 can assess the bilateral upper limb rehabilitation coordination of stroke patients based on the similarity value between the first and second electromyographic signals. The processing module 200 can classify the bilateral upper limb coordination of stroke patients into at least five levels: Level 1, Level 2, Level 3, Level 4, and Level 5, based on the similarity value between the first and second electromyographic signals. Specifically, Level 1: Statistically, the similarity value between the first and second electromyographic signals reaches 91%~100%, indicating good coordination training results during rehabilitation; Level 2: Statistically, the similarity value between the first and second electromyographic signals reaches 91%~100%, indicating good coordination training results during rehabilitation; Level 1: The similarity between the first and second electromyographic signals is 81%–90%, indicating that the patient's coordination training effect needs to be improved during rehabilitation. Level 2: Statistically, the similarity between the first and second electromyographic signals is 61%–80%, indicating that the patient's coordination training effect is poor during rehabilitation. Level 3: Statistically, the similarity between the first and second electromyographic signals is 41%–60%, indicating that the patient's coordination training effect is unsatisfactory during rehabilitation. Level 4: Statistically, the similarity between the first and second electromyographic signals is less than or equal to 40%, indicating that the patient's coordination training effect is extremely poor during rehabilitation.

[0059] In the calculation of the driving torque of the aforementioned training module 300, the driving torque of the training module 300 is configured to increase gradually and slowly, and to stop increasing when the real upper limb shows a tendency to move. The real upper limb is then assisted with driving by adding a surplus value of the current driving torque plus a preset proportion of the current driving torque. However, considering the difference between the first and second electromyographic signals, the calculation of the driving torque applied to both upper limbs needs further improvement. At least, the preset proportion used to calculate the surplus value should be able to be changed based on the magnitude of the difference between the first and second electromyographic signals. Referring to the foregoing, the preset proportion refers to the ratio between the surplus value and the current driving torque, usually expressed and calculated as a fraction of the current driving torque.

[0060] Preferably, the processing module 200 adjusts the magnitude of the surplus value based on the difference between the first electromyographic signal and the second electromyographic signal by adjusting a preset ratio, thereby adjusting the total value of the driving torque applied to one of the upper limbs.

[0061] Specifically, assuming the intensity of the first electromyographic signal in the left upper limb is greater than the intensity of the second electromyographic signal in the right upper limb, it indicates that the recovery of the left upper limb is better than that of the right upper limb. Therefore, when the driving torque is applied by the training module 300, the driving torque applied to the right upper limb is smaller than that applied to the left upper limb. The design logic of this method is that by applying a smaller driving torque to the right upper limb, which has a poorer recovery, the right upper limb receives more exercise. Furthermore, because a smaller driving torque is applied, the stroke patient's attention will be more focused on driving the right upper limb during the movement of the right upper limb, thereby gradually bringing the recovery effect of the right upper limb to the same level as that of the left upper limb. This achieves the effect of maintaining coordination between the two upper limbs during the rehabilitation training process.

[0062] Preferably, the method for determining the preset ratio based on the difference between the first and second electromyographic signals can be a stepwise correspondence. For example, in the aforementioned patient coordination training grading method during rehabilitation training, Level 1 indicates that the preset ratio used to calculate the surplus value in the driving torque applied to the left and right upper limbs remains consistent; Level 2 indicates that the preset ratio used to calculate the surplus value in the driving torque applied to the left upper limb uses the default value (0.1, 10%), while the preset ratio used to calculate the surplus value in the driving torque applied to the right upper limb can be set to 0.08, i.e., 8%; Level 3 indicates that the preset ratio used to calculate the surplus value in the driving torque applied to the left upper limb uses the default value (0.1, 10%), while the preset ratio used to calculate the surplus value in the driving torque applied to the right upper limb can be set to 0.08, i.e., 8%; Level 3 indicates that the preset ratio used to calculate the surplus value in the driving torque applied to the left upper limb uses the default value (0.1, 10%), while the preset ratio used to calculate the surplus value in the driving torque applied to the right upper limb... The preset percentage used to calculate the surplus value in the driving torque can be set to 0.06, or six percent; Level 4 indicates that the preset percentage used to calculate the surplus value in the driving torque applied to the left upper limb uses the default value (0.1, ten percent), while the preset percentage used to calculate the surplus value in the driving torque applied to the right upper limb can be set to 0.04, or four percent; Level 5 indicates that the preset percentage used to calculate the surplus value in the driving torque applied to the left upper limb uses the default value (0.1, ten percent), while the preset percentage used to calculate the surplus value in the driving torque applied to the right upper limb can be set to 0.02, or two percent.

[0063] Example 4 This embodiment is an improvement and supplement to embodiments 1-3, and repeated content will not be repeated.

[0064] This application proposes an upper limb rehabilitation assessment system for stroke patients, which includes at least a data acquisition module 100, a processing module 200, a training module 300, and a display module 400. The training module 300 is roughly in the form of an exoskeleton, and the display module 400 is a regular display screen.

[0065] Preferably, the acquisition module 100 is mainly divided into an electromyography (EMG) acquisition unit 110 and a motion capture unit 120. The EMG acquisition unit 110 and the motion capture unit 120 are configured separately. The EMG acquisition unit 110 and the training module 300 are configured together on the bilateral upper limbs of the stroke patient. Specifically, the electrode pads, arm rings, sensors, etc. of the EMG acquisition unit 110 are held in the middle by the training module 300 and the skin of the upper limbs of the stroke patient. The motion capture unit 120 of the acquisition module 100 is configured in conjunction with the display module 400. Specifically, the camera device of the motion capture unit 120 is configured on the display module 400.

[0066] Preferably, the processing module 200 can be integrated into the display screen of the display module 400 in the form of software, or it can be installed inside the display module 400 in the form of a chip. The processing module 200 can at least receive information from the acquisition module 100, and after processing the collected real upper limb movement and posture information, construct a virtual upper limb in the screen (virtual environment) of the display module 400. The processing module 200 can also judge the coordination of the left and right upper limbs based on the collected electromyographic signals of the real upper limbs, and output the required auxiliary motion parameters to the training module 300 in combination with the difference between the virtual upper limb and the target state in the display module 400 and the coordination of the left and right upper limbs. The auxiliary motion parameters mainly include the driving angle and the driving torque.

[0067] Preferably, Figure 3A simplified structural diagram of the training module 300 of the upper limb rehabilitation assessment system for stroke patients of the present invention is shown. The training module 300 includes at least a first exoskeleton, a second exoskeleton, a third exoskeleton, and a fourth exoskeleton. The first exoskeleton and the second exoskeleton are connected by a first external joint, the second exoskeleton and the third exoskeleton are connected by a second external joint, and the third exoskeleton and the fourth exoskeleton are connected by a third external joint. Specifically, the first exoskeleton is structurally designed according to the shape of the human shoulder and is used to cover the human shoulder. The second exoskeleton is structurally designed according to the shape of the human upper arm and is used to cover the human upper arm. The third exoskeleton is structurally designed according to the shape of the human forearm and is used to cover the human forearm. The fourth exoskeleton is structurally designed according to the shape of the human palm and fingers and is used to cover the human palm and fingers. The first external joint between the first exoskeleton and the second exoskeleton is similar to the human shoulder joint, the second external joint between the second exoskeleton and the third exoskeleton is similar to the human elbow joint, and the third external joint between the third exoskeleton and the fourth exoskeleton is similar to the human wrist joint. In the existing technology, there are relatively mature technologies for designing exoskeletons, rehabilitation robots, etc. The specific driving principle of the training module 300 can refer to the existing technical principles. The advantage of the training module 300 of this application is that it can be controlled by the processing module 200 so that the driving torque it provides is only used to provide a slight auxiliary force for the upper limb movement of the stroke patient, rather than actively driving the upper limb to move. In addition, the training module 300 of this application can also, based on the coordination between the left and right upper limbs analyzed by the processing module 200, adjust the magnitude of the auxiliary driving torque at the left and right upper limbs so that the upper limb with a worse rehabilitation effect receives more training, allowing the stroke patient to focus more on the upper limb with a worse rehabilitation effect, thereby gradually reducing the difference in the degree of rehabilitation between the two upper limbs and making the two upper limbs more coordinated.

[0068] Based on the above approach, this application focuses on the differences in limb movement between the left and right sides of the patient and provides different and targeted training parameters for each side. While improving the training effect and efficiency, it also avoids the situation where the muscles of the left and right sides of the limbs become uncoordinated during the rehabilitation process due to the application of inappropriate training intensity. It solves the problem that the lack of coordination between the two sides of the limbs can hinder the normal life of patients and has a certain positive impact on the quality of life of stroke patients.

[0069] It should be noted that the specific embodiments described above are exemplary, and those skilled in the art can devise various solutions inspired by the disclosure of this invention. These solutions all fall within the scope of this invention and its protection. Those skilled in the art should understand that this specification and its accompanying drawings are illustrative and not intended to limit the scope of the claims. The scope of protection of this invention is defined by the claims and their equivalents.

Claims

1. A motor rehabilitation coordination assessment system based on electromyographic similarity, characterized in that, include: Acquisition module (100) is used to acquire upper limb information of stroke patients; A processing module (200) is used to process the upper limb information collected by the acquisition module (100); Training module (300) is used to assist stroke patients in performing upper limb movements; Display module (400) is used to display the upper limb movement status of stroke patients; The display module (400) shows the difference between the virtual upper limb and the target state to the stroke patient so that the stroke patient can drive the real upper limb to move. The processing module (200) determines the motor ability of the real upper limb based on the difference between the virtual upper limb and the target state, and controls the training module (300) to perform auxiliary driving according to the motor ability; The processing module (200) evaluates the coordination of the bilateral upper limbs of stroke patients based on the similarity between the first electromyography signal and the second electromyography signal; and the processing module (200) adjusts the size of the surplus value by adjusting the preset ratio based on the difference value between the first electromyography signal and the second electromyography signal, thereby dynamically compensating the total value of the driving torque applied to one of the upper limbs. Wherein, the first electromyographic signal and the second electromyographic signal refer to the electromyographic signal of a patient's upper limb under movement conditions acquired by the electromyographic acquisition unit (110) of the acquisition module (100). The first electromyographic signal refers to the electromyographic signal generated by the movement of the left upper limb, and the second electromyographic signal refers to the electromyographic signal generated by the right upper limb. The acquisition module (100) includes an electromyography (EMG) acquisition unit (110) and a motion capture unit (120). The EMG acquisition unit (110) is used to acquire EMG signal information of different motion states and send it to the processing module (200). The processing module (200) constructs the muscle changes of the virtual upper limb in the display module (400) based on the EMG signal and other information acquired by the EMG acquisition unit (110). The motion capture unit (120) is used to acquire the joint coordinates and spatial posture information of the upper limb and send it to the processing module (200). The processing module (200) adjusts the coordinates and posture of the virtual upper limb in the display module (400) based on the changes in joint coordinates and spatial posture information.

2. The sports rehabilitation coordination assessment system according to claim 1, characterized in that, The processing module (200) classifies the bilateral upper limb coordination of stroke patients into five levels: Level 1, Level 2, Level 3, Level 4, and Level 5, based on the similarity value between the first and second electromyographic signals. Level 1 is defined as a similarity value of 91% to 100% between the first and second electromyographic signals; Level 2 is defined as a similarity value of 81% to 90% between the first and second electromyographic signals; Level 3 is defined as a similarity value of 61% to 80% between the first and second electromyographic signals; Level 4 is defined as a similarity value of 41% to 60% between the first and second electromyographic signals; and Level 5 is defined as a similarity value of less than or equal to 40% between the first and second electromyographic signals.

3. The sports rehabilitation coordination assessment system according to claim 2, characterized in that, The processing module (200) controls the driving torque of the training module (300) to gradually increase, and the training module (300) stops increasing when the real upper limb has a movement tendency, and provides auxiliary driving to the real upper limb with the current driving torque plus a surplus value of the current driving torque preset ratio.

4. The sports rehabilitation coordination assessment system according to claim 3, characterized in that, The method for determining the preset ratio based on the difference between the first and second electromyographic signals is a step-by-step correspondence: Level 1 indicates that the preset ratio used to calculate the surplus value in the driving torque applied to the left and right upper limbs remains consistent; Level 2 indicates that the preset ratio used to calculate the surplus value in the driving torque applied to the left upper limb uses the default value, while the preset ratio used to calculate the surplus value in the driving torque applied to the right upper limb is set to 0.08; Level 3 indicates that the preset ratio used to calculate the surplus value in the driving torque applied to the left upper limb uses the default value, while the preset ratio used to calculate the surplus value in the driving torque applied to the right upper limb is set to 0.06; Level 4 indicates that the preset ratio used to calculate the surplus value in the driving torque applied to the left upper limb uses the default value, while the preset ratio used to calculate the surplus value in the driving torque applied to the right upper limb is set to 0.04; Level 5 indicates that the preset ratio used to calculate the surplus value in the driving torque applied to the left upper limb uses the default value, while the preset ratio used to calculate the surplus value in the driving torque applied to the right upper limb is set to 0.

02.

5. The sports rehabilitation coordination assessment system according to claim 1, characterized in that, The processing module (200) can analyze the time and motion state of the virtual upper limb in the display module (400) from moving to the target state, and judge the motion ability of the real upper limb in the actual environment based on the time and motion state. When the processing module (200) believes that the motion ability of the real upper limb is insufficient to complete the motion from the initial state to the target state, the processing module (200) analyzes the driving angle and driving torque required for the training module (300) to perform auxiliary driving based on inverse kinematics analysis. The driving torque calculated by the processing module (200) is the maximum driving torque that the training module (300) can provide under the current auxiliary driving.

6. The sports rehabilitation coordination assessment system according to claim 5, characterized in that, The processing module (200) calculates the driving angle and driving torque by: obtaining the current position of the virtual upper limb end and the target position of the target state, and calculating the driving angle of each joint based on the current position and the target position using an inverse kinematics algorithm; dynamically adjusting the arm support torque of each joint based on iterative learning, and calculating the arm support torque required for each joint based on the target driving angle and the actual joint angle, target joint angular velocity, and actual joint angular velocity; performing inverse dynamics control based on feedback linearization, calculating the output torque of each joint based on the target motion parameters and actual motion parameters of each joint, and calculating the driving amount of each joint based on the output torque of each joint.

7. The sports rehabilitation coordination assessment system according to claim 1, characterized in that, The processing module (200) is at least able to analyze the electromyographic signals acquired by the acquisition module (100) to determine the muscle contraction state, and determine whether the muscle movement is normal according to the preset muscle contraction threshold; the processing module (200) monitors the abnormal muscle contraction state, and determines whether the muscle is in an abnormal movement state by calculating the difference between the muscle contraction value obtained from the real-time acquired electromyographic signals and the preset abnormal muscle contraction threshold. The abnormal muscle contraction mode includes muscle overexcitation, muscle overinhibition, abnormal muscle coordination, abnormal co-contraction of agonist and antagonist muscles, and muscle fatigue.

8. The sports rehabilitation coordination assessment system according to claim 1, characterized in that, The processing module (200) maps the real upper limb to the virtual upper limb in the virtual environment. The steps include: building a model of the virtual upper limb in the virtual environment; collecting electromyographic signals and posture information of the real upper limb; extracting features from the electromyographic signals, motion information features and posture information of the real upper limb, and classifying them in a hierarchical manner according to the strength of the features; fusing the electromyographic features and the angle features of the motion and inputting them into a classifier for pattern recognition; dynamically recognizing the actions and movements of the real upper limb and synchronously adjusting the actions and movements of the virtual upper limb.

9. The sports rehabilitation coordination assessment system according to claim 1, characterized in that, The electromyography acquisition unit (110) employs a surface electromyography sensor, electromyography electrodes, and / or an electromyography arm ring; the electromyography arm ring is worn on the forearm and upper arm of the stroke patient to collect bioelectric signals generated by the muscles of the forearm and upper arm, motion acceleration and displacement signals generated by movement, respectively.

10. The sports rehabilitation coordination assessment system according to claim 9, characterized in that, The processing module (200) analyzes the current movement of the upper limb based on the motion capture unit (120), and judges the vitality of each muscle in the rehabilitation process based on the electromyographic signals of the corresponding muscles among the deltoid, biceps brachii, triceps brachii, brachialis, pronator teres, brachioradialis, extensor carpi radialis longus, extensor carpi radialis brevis, flexor carpi radialis, palmaris longus, flexor digitorum superficialis, and extensor carpi ulnaris actually collected by the electromyographic acquisition unit (110).

11. The sports rehabilitation coordination assessment system according to claim 1, characterized in that, The motor rehabilitation coordination assessment system is configured as a portable home system; the stroke patient imagines himself performing the same movement through the motor imagery displayed by the display module (400), and the processing module (200) evaluates the rehabilitation effect and the coordination between the upper limbs of the stroke patient in the individual family by analyzing and comparing the electromyographic signals generated by the upper limb movement.

Images