A collaborative rehabilitation method, system, terminal, and storage medium based on upper limb exoskeleton task-driven approach for healthy and affected sides.

Abstract

This invention discloses a method, system, terminal, and storage medium for collaborative rehabilitation of the healthy and affected sides based on task-driven upper limb exoskeleton. The method includes: having the target patient wear an exoskeleton and a surface-mounted electromyography (EMG) sensor; using the movement characteristics of the healthy arm, providing different task scenarios, and learning the movement trajectory of the affected side with the goal of completing the task; simultaneously performing rehabilitation rating based on the EMG signals; assigning impedance parameters for the interaction between the exoskeleton and the arm based on the rating; and also estimating the movement intention through a muscle model and compensating and correcting the imitated movement trajectory in real time. This allows the patient to achieve collaboration between the healthy and affected sides during task completion, promoting functional recovery of the damaged limb. Through repeated practice of asymmetrical movements, the patient's motor control ability and bilateral limb coordination ability are enhanced during later rehabilitation, restoring basic motor function of the damaged limb, making the rehabilitation training process more consistent with the patient's rehabilitation process, and improving rehabilitation effectiveness.

Description

Technical Field

[0001] This invention relates to the field of exoskeleton robot rehabilitation technology, and in particular to a method, system, terminal, and computer-readable storage medium for collaborative rehabilitation of the healthy and affected sides based on task-driven upper limb exoskeleton. Background Technology

[0002] Stroke, commonly known as a "cerebrovascular accident," is a persistent loss of brain function caused by acute cerebrovascular disease. Stroke patients often experience varying degrees of functional impairment after rehabilitation, affecting multiple areas including motor, sensory, cognitive, language, and psychological aspects. Upper limb dysfunction, in particular, is very common among stroke and traumatic brain injury patients.

[0003] In the field of rehabilitation medicine, traditional stroke rehabilitation typically involves one-on-one direct contact between a physical therapist and the patient. This involves prolonged limb-assisted training, combined with verbal communication and medication, to help patients perform numerous repetitive movements to stimulate damaged brain nerves and promote repair. However, this traditional rehabilitation method is inefficient, costly, and difficult to perform at home. In recent years, with technological advancements, upper limb exoskeleton robots have emerged as a mechanical system suitable for prolonged, repetitive labor. These robots can provide high-intensity, highly stable repetitive training while ensuring efficiency. With a structure similar to the human upper limb, they can provide gravity compensation for the affected limb and possess multiple degrees of freedom, assisting patients in performing complex upper limb joint movements. Using upper limb exoskeleton robots can significantly improve the limitations of traditional rehabilitation methods, providing stroke patients with a more efficient, economical, and convenient rehabilitation solution.

[0004] For existing stroke patients, a relatively mature method is mirror image therapy, which combines the patient's healthy side motor signals with the control strategy of the assistive device on the affected side. It integrates active rehabilitation medicine and robotics, and the bilateral symmetrical upper limb motor pattern based on the theory of neuroplasticity enables the undamaged hemisphere and the damaged hemisphere to interact, which helps the nervous system recover and the rehabilitation of motor disorders. It replaces traditional rehabilitation therapists and provides a new approach for the upper limb function rehabilitation of patients with motor disorders.

[0005] The drawbacks of existing technologies are as follows: First, their applicability is limited, mainly applicable to patients with unilateral limb injuries, and may not be suitable for patients with bilateral injuries. Second, precise synchronization control technology is required to ensure that the movements of the healthy and injured limbs are consistent. Third, parameter adjustments are complex and require individualized adjustments based on the patient's specific situation to ensure optimal rehabilitation results. Finally, there is the potential for over-reliance; if patients become overly dependent on the synchronized movements of the healthy limb, it may weaken the independent movement ability of the injured limb, and some basic movements requiring asymmetrical actions may never be recovered.

[0006] Therefore, existing technologies still need to be improved and developed. Summary of the Invention

[0007] The main objective of this invention is to provide a collaborative rehabilitation method, system, terminal, and computer-readable storage medium based on upper limb exoskeleton task-driven approach for healthy and affected sides, aiming to address the significant limitations and poor rehabilitation outcomes of existing mirror rehabilitation methods for stroke patients.

[0008] To achieve the above objectives, the present invention provides a collaborative rehabilitation method for the healthy and affected sides based on task-driven upper limb exoskeleton. This method includes the following steps:

[0009] Obtain the patient's disease information, assess the upper limb functional status of the target patient based on the patient's disease information, match the upper limb exoskeleton device to be worn based on the characteristics of the target patient, and wear a surface-mounted electromyography sensor;

[0010] Surface electromyography (EMG) signals from the target muscle group on the healthy side of the target patient are collected, key features of the EMG signals are extracted, motion data of the healthy arm dragging the upper limb exoskeleton device are captured, a muscle model is established, and motion intention is estimated based on the key features of the EMG signals and the motion data using the muscle model.

[0011] The position sensor data on the upper limb exoskeleton device is input into the DMP model to complete the imitation learning of the affected side, and the trained DMP model is used to generate a motion trajectory that matches the patient's actual condition.

[0012] Multiple specific bi-arm collaborative rehabilitation tasks are set, and each rehabilitation task involving the healthy and affected sides is completed through the upper limb exoskeleton device. Evaluation results are obtained based on the completion rate of each rehabilitation task.

[0013] Optionally, the above-described task-driven collaborative rehabilitation method for the healthy and affected sides based on an upper limb exoskeleton, wherein acquiring the target patient's disease information, assessing the target patient's upper limb functional status based on the disease information, matching the required upper limb exoskeleton device based on the target patient's characteristics, and wearing a surface-mounted electromyography sensor, specifically includes:

[0014] Obtain the target patient's disease information and assess the target patient's upper limb functional status based on the patient's disease information using the Fugl-Meyer scale, a clinical assessment tool.

[0015] Based on the characteristics of the target patient, a size-matched upper limb exoskeleton device is determined. The upper limb exoskeleton device is used to cover the upper limb joints on the affected and unaffected sides of the target patient, and surface-mounted electromyography (EMG) sensors are worn at designated key muscle locations on the unaffected and affected sides of the target patient.

[0016] Optionally, the above-described task-driven collaborative rehabilitation method based on an upper limb exoskeleton, wherein the step of collecting surface electromyography (EMG) signals from the target muscle groups on the healthy side of the target patient, extracting key features of the EMG signals, capturing motion data of the healthy arm dragging the upper limb exoskeleton device, establishing a muscle model, and estimating the movement intention based on the key features of the EMG signals and the motion data using the muscle model, specifically includes:

[0017] Surface electromyography (EMG) signals from the target muscle group on the healthy side of the target patient are collected. Noise in the surface EMG signals is removed using a digital filter, and key features of the surface EMG signals are extracted and saved.

[0018] The target patient is instructed to use their unaffected arm to drag the upper limb exoskeleton device and move it, and the motion data during the movement is recorded. The motion data includes the position of the joints, torque, and six-dimensional force values ​​at the upper limb interaction points.

[0019] In the OpenSim environment, the joint positions, torques, and six-dimensional force values ​​of the upper limb interaction points captured by the motion capture of the dragging upper limb exoskeleton device are imported. The joint torques are derived using the inverse dynamics model of the upper limb exoskeleton, the torques generated by muscle activity are estimated, and the parameters are adjusted to make the model output match the experimental data.

[0020] Establish a muscle model and optimize the muscle parameters of the muscle model, including optimal fiber length, tendon relaxation length, muscle belly angle, default activation, default fiber length, and contraction speed coefficient.

[0021] The muscle model estimates the movement intention based on the key features of the surface electromyography signal, and analyzes the specific action that the target patient wants to perform.

[0022] Optionally, in the above-mentioned task-driven collaborative rehabilitation method based on upper limb exoskeleton, the muscle model includes an SVM classification model;

[0023] The muscle model estimates motor intent based on key features of the surface electromyography signals, specifically including:

[0024] During the training phase, an SVM classification model is constructed using known motion actions and corresponding EMG features.

[0025] During the testing phase, new EMG data is introduced to evaluate the accuracy and generalization ability of the SVM classification model, ensuring that the SVM classification model works correctly on the new data.

[0026] After model training and validation are completed, the trained SVM classification model is used to predict the target patient's movement intention based on the key features of the surface electromyography signal.

[0027] Optionally, the above-described task-driven collaborative rehabilitation method for the healthy and affected sides based on an upper limb exoskeleton, wherein the step of inputting position sensor data from the upper limb exoskeleton device into a DMP model to complete the affected side's imitation learning, and using the trained DMP model to generate a motion trajectory that conforms to the patient's actual condition, specifically includes:

[0028] The position sensor data from the upper limb exoskeleton device is input into the DMP model, and a Gaussian function is selected as the basis function to determine the number, center position, and width of the basis function.

[0029] The weights of the basis functions are calculated using the least squares method, so that the generated motion trajectory is consistent with the actual motion trajectory of the target patient.

[0030] Adjust the natural frequency parameters to control the rhythm of the movement, and adjust the impedance coefficient to control the contact force between the target patient and the exoskeleton.

[0031] During the DMP model training process, the motion state of the target patient is monitored in real time through the joint encoder on the upper limb exoskeleton device and the forward and inverse kinematics model of the exoskeleton. The DMP parameters are dynamically adjusted according to the real-time data to ensure the accuracy and smoothness of the motion trajectory. The trained DMP model is used to generate a motion trajectory that conforms to the actual condition of the target patient.

[0032] Optionally, the above-described task-driven healthy-affected side collaborative rehabilitation method based on an upper limb exoskeleton includes setting multiple specific bi-arm collaborative rehabilitation tasks, completing each healthy-affected side collaborative rehabilitation task through the upper limb exoskeleton device, and obtaining evaluation results based on the completion degree of each rehabilitation task, specifically including:

[0033] Set up multiple specific bi-arm collaborative rehabilitation tasks and ensure that each rehabilitation task is relevant to the daily life of the target patient;

[0034] The upper limb exoskeleton device is used to complete rehabilitation tasks in collaboration between the healthy and affected sides, determine the completion rate of each rehabilitation task, and obtain evaluation results based on the completion rate of each rehabilitation task.

[0035] Optionally, the above-described task-driven collaborative rehabilitation method based on an upper limb exoskeleton, wherein the evaluation result is obtained based on the completion of each rehabilitation task, further includes:

[0036] The parameters of the DMP model and the rehabilitation training program are adjusted based on the evaluation results to achieve personalized rehabilitation training.

[0037] Furthermore, to achieve the above objectives, the present invention also provides a collaborative rehabilitation system for the healthy and affected sides based on a task-driven upper limb exoskeleton, wherein the collaborative rehabilitation system for the healthy and affected sides based on a task-driven upper limb exoskeleton includes:

[0038] The assessment and matching module is used to acquire the patient's disease information, assess the upper limb functional status of the target patient based on the patient's disease information, match the upper limb exoskeleton device to be worn based on the characteristics of the target patient, and wear a surface-mount electromyography sensor.

[0039] The acquisition and estimation module is used to acquire surface electromyography (EMG) signals from the target muscle group on the healthy side of the target patient, extract key features of the surface EMG signals, capture motion data of the healthy arm dragging the upper limb exoskeleton device, establish a muscle model, and estimate the motion intention based on the key features of the surface EMG signals and the motion data through the muscle model.

[0040] The learning and generation module is used to input the position sensor data of the upper limb exoskeleton device into the DMP model to complete the imitation learning of the affected side, and use the trained DMP model to generate a motion trajectory that conforms to the patient's actual condition.

[0041] The rehabilitation and assessment module is used to set multiple specific bi-arm collaborative rehabilitation tasks. Each rehabilitation task involving the healthy and affected sides is completed through the upper limb exoskeleton device, and assessment results are obtained based on the completion rate of each rehabilitation task.

[0042] In addition, to achieve the above objectives, the present invention also provides a terminal, wherein the terminal includes: a memory, a processor, and a healthy-affected side collaborative rehabilitation program based on upper limb exoskeleton task-driven method stored in the memory and executable on the processor. When the healthy-affected side collaborative rehabilitation program based on upper limb exoskeleton task-driven method is executed by the processor, it implements the steps of the healthy-affected side collaborative rehabilitation method based on upper limb exoskeleton task-driven method as described above.

[0043] Furthermore, to achieve the above objectives, the present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a healthy-affected side collaborative rehabilitation program based on upper limb exoskeleton task-driven method, wherein when the healthy-affected side collaborative rehabilitation program based on upper limb exoskeleton task-driven method is executed by a processor, it implements the steps of the healthy-affected side collaborative rehabilitation method based on upper limb exoskeleton task-driven method as described above.

[0044] In this invention, patient disease information of the target patient is acquired, and the upper limb functional status of the target patient is assessed based on the patient's disease information. An upper limb exoskeleton device is matched to the target patient's characteristics, and a surface-mounted electromyography (EMG) sensor is worn. Surface EMG signals from the target muscle groups on the unaffected side of the target patient are collected, key features of the surface EMG signals are extracted, and motion data of the unaffected arm dragging the upper limb exoskeleton device is captured to establish a muscle model. The muscle model is used to estimate the movement intention based on the key features of the surface EMG signals and the motion data. Position sensor data from the upper limb exoskeleton device is input into a DMP model to complete the affected side's imitation learning, and the trained DMP model is used to generate a movement trajectory that matches the patient's actual condition. Multiple specific bi-arm collaborative rehabilitation tasks are set, and each rehabilitation task involving the unaffected and affected sides is completed using the upper limb exoskeleton device. Evaluation results are obtained based on the completion rate of each rehabilitation task. This invention enables patients to coordinate their healthy and affected sides while completing tasks, promoting the functional recovery of the damaged limbs. Through repeated practice of asymmetrical movements, it enhances the patient's motor control and bilateral limb coordination during later rehabilitation, ultimately restoring the basic motor functions of the damaged limbs. This makes the rehabilitation training process more aligned with the patient's rehabilitation process, improving rehabilitation outcomes while satisfying human-computer interaction requirements. Attached Figure Description

[0045] Figure 1 This is a flowchart of a preferred embodiment of the present invention, which is a collaborative rehabilitation method for the healthy and affected sides based on a task-driven upper limb exoskeleton.

[0046] Figure 2 This is a schematic diagram of the exoskeleton rehabilitation robot structure in a preferred embodiment of the present invention, which is a collaborative rehabilitation method for the healthy and affected sides based on task-driven upper limb exoskeleton.

[0047] Figure 3 This is a schematic diagram of the control process in a preferred embodiment of the healthy-affected side collaborative rehabilitation method based on upper limb exoskeleton task-driven approach of the present invention.

[0048] Figure 4 This is a structural diagram of a preferred embodiment of the healthy-affected side collaborative rehabilitation system based on upper limb exoskeleton task-driven design of the present invention;

[0049] Figure 5 This is a structural diagram of a preferred embodiment of the terminal of the present invention. Detailed Implementation

[0050] To make the objectives, technical solutions, and advantages of this invention clearer and more explicit, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0051] The preferred embodiment of the present invention describes a collaborative rehabilitation method for the affected and healthy sides based on a task-driven upper limb exoskeleton, such as... Figure 1 As shown, the collaborative rehabilitation method for the healthy and affected sides based on upper limb exoskeleton task-driven approaches includes the following steps:

[0052] Step S10: Obtain the patient's disease information, assess the upper limb functional status of the target patient based on the patient's disease information, match the upper limb exoskeleton device to be worn based on the characteristics of the target patient, and wear a surface-mounted electromyography sensor.

[0053] Specifically, the patient's disease information (such as the cause of onset, time of onset, treatment history, etc., to facilitate a detailed understanding of the patient's medical history; the patient's disease information is only used for data analysis and does not involve the patient's privacy information, and is strictly confidential and prohibited from being disclosed) is obtained. The upper limb functional status of the target patient is assessed based on the patient's disease information using the Fugl-Meyer scale, a clinical assessment tool. The muscle activity status of the patient's upper limb is examined by electromyography to understand the excitability and coordination of the muscles.

[0054] Then, based on the characteristics of the target patient, a matching upper limb exoskeleton device is determined (ensuring the device size is appropriate). This exoskeleton device covers the affected and unaffected upper limb joints of the target patient. Surface-attached electromyography (SEMG) sensors are worn at designated key muscle locations (e.g., biceps brachii, triceps brachii, forearm extensor muscles) on both the affected and unaffected sides of the target patient, ensuring close contact between the SEMG sensors and the skin to avoid signal interference. The upper limb exoskeleton device is then adjusted to wear mode, and the target patient is guided to wear the exoskeleton device correctly, ensuring a good fit between the exoskeleton device and the target patient's body.

[0055] like Figure 2 The diagram shows the structure of an exoskeleton rehabilitation robot (i.e., an upper limb exoskeleton device). 1-1 represents the right arm of the upper limb rehabilitation exoskeleton; 1-2 represents the left arm of the upper limb rehabilitation exoskeleton; 2-1 represents the six-dimensional force sensor at the interaction point of the right upper arm; 2-2 represents the six-dimensional force sensor at the interaction point of the left upper arm; 3 represents a surface electromyography (SEMG) sensor wristband (which collects the patient's SEMG signals for movement intention recognition); 4-1 represents the six-dimensional force sensor at the interaction point of the right forearm (the six-dimensional force sensor is mounted on the exoskeleton and is used to detect the interaction force with the patient in real time); 4-2 represents the six-dimensional force sensor at the interaction point of the left forearm; and 5 represents the task implementation prop.

[0056] Step S20: Collect surface electromyography (EMG) signals from the target muscle group on the healthy side of the target patient, extract key features of the EMG signals, capture motion data of the healthy arm dragging the upper limb exoskeleton device, establish a muscle model, and estimate the movement intention based on the key features of the EMG signals and the motion data using the muscle model.

[0057] Specifically, high-quality surface electromyography (EMG) signals from the target muscle group on the healthy side of the target patient are collected (as raw data). The collected EMG signals are processed, and a digital filter is used to remove noise from the EMG signals. Key features of the EMG signals are extracted and saved (i.e., the processed EMG data is saved). Then, the target patient is instructed to use the healthy arm to drag the upper limb exoskeleton device and the motion data during the movement is recorded. The motion data includes joint positions, torques, and six-dimensional force values ​​at the upper limb interaction points. These data are crucial for constructing an accurate inverse dynamics model because they provide information about how the limb responds to external forces.

[0058] In the OpenSim environment, the positions, torques, and six-dimensional force values ​​of the joints captured by the motion capture of the upper limb exoskeleton device are imported. The joint torques are derived using the inverse dynamics model of the upper limb exoskeleton, the torques generated by muscle activity are estimated, and the parameters are adjusted to make the model output match the experimental data. This involves calculating the torque at each joint and then inferring which muscles are responsible for the specific movements.

[0059] A muscle model is established, and the muscle parameters of the muscle model are optimized. The muscle parameters include optimal fiber length, tendon relaxation length, muscle belly angle, default activation, default fiber length, and contraction speed coefficient, which can make the simulation results closer to the actual situation. Data collected with the assistance of an exoskeleton is used to calibrate the muscle model to ensure that it can truly reflect actual body movements.

[0060] The muscle model estimates motor intent based on key features of the surface electromyography (EMG) signals, thus deciphering the specific action the target patient intends to perform. The muscle model includes an SVM classification model. During the motor intent estimation process using the muscle model, in the training phase, an SVM classification model is constructed using known motor actions and corresponding EMG features. In the testing phase, new EMG data is introduced to evaluate the accuracy and generalization ability of the SVM classification model, ensuring it functions correctly on the new data. After model training and validation are complete, the trained SVM classification model is used to predict the target patient's motor intent based on the key features of the surface EMG signals. This process involves mapping EMG features to specific outputs, which can be discrete action categories, such as grasping, clenching a fist, or opening a palm, or continuous motion parameters. In this way, the specific action the user wants to perform can be deciphered.

[0061] Step S30: Input the position sensor data of the upper limb exoskeleton device into the DMP model to complete the imitation learning of the affected side, and use the trained DMP model to generate a motion trajectory that matches the patient's actual condition.

[0062] Specifically, such as Figure 3 As shown, during the affected side imitation learning process, the position sensor data from the upper limb exoskeleton device is input into the DMP (Dynamic Movement Primitives) model (the DMP model's function is to mimic the movement characteristics of the healthy side to generate the movement trajectory of the affected side). A Gaussian function is selected as the basis function, and the number, center position, and width of the basis functions are determined. The weights of the basis functions are calculated using the least squares method (in the DMP framework, the movement trajectory is represented by a linear combination of a set of basis functions, each of which describes a local characteristic of the movement trajectory. The least squares method calculates the weights of the basis functions to minimize the difference between the generated movement trajectory and the actual movement trajectory), so that the generated movement trajectory is consistent with the actual movement trajectory of the target patient. During this process, the natural frequency parameter can be adjusted (in the DMP model, the natural frequency parameter affects the time scale or period of the generated movement trajectory. By adjusting the natural frequency parameter, the performance of the DMP can be optimized to better match the required task characteristics and personal usage habits) to control the rhythm of the movement, and the impedance coefficient can be adjusted to control the contact force between the target patient and the exoskeleton.

[0063] During the DMP model training process, the movement status of the target patient is monitored in real time through the joint encoder on the upper limb exoskeleton device and the forward and inverse kinematics model of the exoskeleton. The DMP parameters (such as natural frequency and different target positions for different tasks) are dynamically adjusted according to the real-time data to ensure the accuracy and smoothness of the movement trajectory. The trained DMP model is used to generate a movement trajectory that conforms to the actual condition of the target patient.

[0064] Step S40: Set multiple specific bi-arm collaborative rehabilitation tasks, complete each rehabilitation task involving the healthy side and the affected side using the upper limb exoskeleton device, and obtain evaluation results based on the completion rate of each rehabilitation task.

[0065] Specifically, several specific bi-arm collaborative rehabilitation tasks are set, such as picking up objects and pouring water, and each rehabilitation task is ensured to be relevant to the target patient's daily life (e.g., ...). Figure 2 (The action of pushing a box with both hands); the rehabilitation task of each healthy side and the affected side is completed through the upper limb exoskeleton device, and the completion degree of each rehabilitation task is determined. The completion degree is related to the score of the sub-tasks into which the task is divided. Completing a certain number of sub-tasks will result in a corresponding score, which is regarded as the completion degree.

[0066] For example, grasping and releasing objects:

[0067] Grasp: to grasp an object with the fingers.

[0068] Release: Let go of the object you are holding.

[0069] For example, moving the arm and hand:

[0070] Raise your arm: Raise your arm from one side of your body to the other side.

[0071] Lower your arm: Put your raised arm back into place.

[0072] Horizontal movement: Moving the arm on a horizontal plane.

[0073] Vertical movement: Moving the arm in the vertical direction.

[0074] Finally, the patient is assisted in completing tasks involving the healthy and affected sides using the exoskeleton device. Evaluation results are obtained based on the completion rate of each rehabilitation task.

[0075] With the assistance of an exoskeleton, patients can complete tasks by coordinating the healthy and affected sides, promoting the functional recovery of the damaged limbs and enabling patients to better regain their motor abilities.

[0076] Furthermore, the parameters of the DMP model and the rehabilitation training program are adjusted based on the assessment results to achieve personalized rehabilitation training. The difficulty of rehabilitation tasks is adjusted, gradually increasing in complexity to accommodate the patient's improved abilities.

[0077] This invention addresses the limitations of mirror-image rehabilitation in the later stages. In this method, the patient wears an exoskeleton and a surface-mounted electromyography (SEMG) sensor. Based on the movement characteristics of the healthy arm, different task scenarios are provided, with the goal of completing the task. Motion primitives (DMP) technology is used to mimic and learn the movement trajectory of the affected side. Simultaneously, the SEMG sensor performs rehabilitation assessment based on the electromyographic signals. Based on the assessment, impedance parameters for the interaction between the exoskeleton and the arm are assigned. Muscle models can also estimate movement intentions and compensate for and correct the mimicked movement trajectory in real time. This allows the patient to achieve collaboration between the healthy and affected sides during task completion, promoting functional recovery of the damaged limb. Through repeated practice of asymmetrical movements, the patient's motor control and bilateral limb coordination are enhanced in the later stages of rehabilitation, ultimately restoring basic motor function of the damaged limb. This invention makes the rehabilitation training process more aligned with the patient's rehabilitation process, improving rehabilitation outcomes while satisfying human-computer interaction requirements.

[0078] Furthermore, such as Figure 4 As shown, based on the above-mentioned upper limb exoskeleton task-driven collaborative rehabilitation method for the healthy and affected sides, the present invention also provides a corresponding upper limb exoskeleton task-driven collaborative rehabilitation system for the healthy and affected sides, wherein the upper limb exoskeleton task-driven collaborative rehabilitation system for the healthy and affected sides includes:

[0079] The assessment and matching module 51 is used to acquire the patient's disease information of the target patient, assess the upper limb functional status of the target patient based on the patient's disease information, match the upper limb exoskeleton device to be worn based on the characteristics of the target patient, and wear the surface-mount electromyography sensor.

[0080] The acquisition and estimation module 52 is used to acquire surface electromyography (EMG) signals from the target muscle group on the healthy side of the target patient, extract key features of the surface EMG signals, capture motion data of the healthy arm dragging the upper limb exoskeleton device, establish a muscle model, and estimate the motion intention based on the key features of the surface EMG signals and the motion data through the muscle model.

[0081] The learning and generation module 53 is used to input the position sensor data of the upper limb exoskeleton device into the DMP model to complete the imitation learning of the affected side, and use the trained DMP model to generate a motion trajectory that conforms to the actual condition of the patient.

[0082] The rehabilitation and assessment module 54 is used to set multiple specific bi-arm collaborative rehabilitation tasks. Each rehabilitation task involving the healthy side and the affected side is completed through the upper limb exoskeleton device, and the assessment results are obtained based on the completion degree of each rehabilitation task.

[0083] Furthermore, such as Figure 5 As shown, based on the above-mentioned collaborative rehabilitation method and system for healthy and affected sides driven by upper limb exoskeleton, the present invention also provides a terminal, which includes a processor 10, a memory 20 and a display 30. Figure 5 Only some of the terminal components are shown; however, it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.

[0084] In some embodiments, the memory 20 may be an internal storage unit of the terminal, such as a hard disk or memory. In other embodiments, the memory 20 may be an external storage device of the terminal, such as a plug-in hard disk, smart media card (SMC), secure digital card (SD), flash card, etc. Further, the memory 20 may include both internal and external storage devices. The memory 20 is used to store application software and various types of data installed on the terminal, such as program code installed on the terminal. The memory 20 can also be used to temporarily store data that has been output or will be output. In one embodiment, the memory 20 stores a task-driven, healthy-affected side collaborative rehabilitation program 40 based on an upper limb exoskeleton. This task-driven, healthy-affected side collaborative rehabilitation program 40 can be executed by the processor 10, thereby implementing the task-driven, healthy-affected side collaborative rehabilitation method based on an upper limb exoskeleton as described in this application.

[0085] In some embodiments, the processor 10 may be a central processing unit (CPU), a microprocessor, or other data processing chip, used to run program code stored in the memory 20 or process data, such as executing the healthy-affected side collaborative rehabilitation method based on upper limb exoskeleton task-driven method.

[0086] In some embodiments, the display 30 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, or an OLED (Organic Light-Emitting Diode) touchscreen. The display 30 is used to display information on the terminal and to display a visual user interface. The terminal's processor 10, memory 20, and display 30 communicate with each other via a system bus.

[0087] In one embodiment, when the processor 10 executes the healthy-affected side collaborative rehabilitation program 40 based on upper limb exoskeleton task-driven architecture stored in the memory 20, the following steps are performed:

[0088] Obtain the patient's disease information, assess the upper limb functional status of the target patient based on the patient's disease information, match the upper limb exoskeleton device to be worn based on the characteristics of the target patient, and wear a surface-mounted electromyography sensor;

[0089] Surface electromyography (EMG) signals from the target muscle group on the healthy side of the target patient are collected, key features of the EMG signals are extracted, motion data of the healthy arm dragging the upper limb exoskeleton device are captured, a muscle model is established, and motion intention is estimated based on the key features of the EMG signals and the motion data using the muscle model.

[0090] The position sensor data on the upper limb exoskeleton device is input into the DMP model to complete the imitation learning of the affected side, and the trained DMP model is used to generate a motion trajectory that matches the patient's actual condition.

[0091] Multiple specific bi-arm collaborative rehabilitation tasks are set, and each rehabilitation task involving the healthy and affected sides is completed through the upper limb exoskeleton device. Evaluation results are obtained based on the completion rate of each rehabilitation task.

[0092] The steps of acquiring the target patient's disease information, assessing the target patient's upper limb functional status based on the disease information, matching the required upper limb exoskeleton device based on the target patient's characteristics, and wearing a surface-mounted electromyography (EMG) sensor specifically include:

[0093] Obtain the target patient's disease information and assess the target patient's upper limb functional status based on the patient's disease information using the Fugl-Meyer scale, a clinical assessment tool.

[0094] Based on the characteristics of the target patient, a size-matched upper limb exoskeleton device is determined. The upper limb exoskeleton device is used to cover the upper limb joints on the affected and unaffected sides of the target patient, and surface-mounted electromyography (EMG) sensors are worn at designated key muscle locations on the unaffected and affected sides of the target patient.

[0095] Specifically, the process involves collecting surface electromyography (EMG) signals from the target muscle groups on the healthy side of the target patient, extracting key features from the EMG signals, capturing motion data of the healthy arm dragging the upper limb exoskeleton device, establishing a muscle model, and estimating the movement intention based on the key features of the EMG signals and the motion data using the muscle model.

[0096] Surface electromyography (EMG) signals from the target muscle group on the healthy side of the target patient are collected. Noise in the surface EMG signals is removed using a digital filter, and key features of the surface EMG signals are extracted and saved.

[0097] The target patient is instructed to use their unaffected arm to drag the upper limb exoskeleton device and move it, and the motion data during the movement is recorded. The motion data includes the position of the joints, torque, and six-dimensional force values ​​at the upper limb interaction points.

[0098] In the OpenSim environment, the joint positions, torques, and six-dimensional force values ​​of the upper limb interaction points captured by the motion capture of the dragging upper limb exoskeleton device are imported. The joint torques are derived using the inverse dynamics model of the upper limb exoskeleton, the torques generated by muscle activity are estimated, and the parameters are adjusted to make the model output match the experimental data.

[0099] Establish a muscle model and optimize the muscle parameters of the muscle model, including optimal fiber length, tendon relaxation length, muscle belly angle, default activation, default fiber length, and contraction speed coefficient.

[0100] The muscle model estimates the movement intention based on the key features of the surface electromyography signal, and analyzes the specific action that the target patient wants to perform.

[0101] The muscle model includes an SVM classification model;

[0102] The muscle model estimates motor intent based on key features of the surface electromyography signals, specifically including:

[0103] During the training phase, an SVM classification model is constructed using known motion actions and corresponding EMG features.

[0104] During the testing phase, new EMG data is introduced to evaluate the accuracy and generalization ability of the SVM classification model, ensuring that the SVM classification model works correctly on the new data.

[0105] After model training and validation are completed, the trained SVM classification model is used to predict the target patient's movement intention based on the key features of the surface electromyography signal.

[0106] Specifically, the step of inputting position sensor data from the upper limb exoskeleton device into the DMP model to complete the affected side's imitation learning, and using the trained DMP model to generate a motion trajectory that matches the patient's actual condition, includes:

[0107] The position sensor data from the upper limb exoskeleton device is input into the DMP model, and a Gaussian function is selected as the basis function to determine the number, center position, and width of the basis function.

[0108] The weights of the basis functions are calculated using the least squares method, so that the generated motion trajectory is consistent with the actual motion trajectory of the target patient.

[0109] Adjust the natural frequency parameters to control the rhythm of the movement, and adjust the impedance coefficient to control the contact force between the target patient and the exoskeleton.

[0110] During the DMP model training process, the motion state of the target patient is monitored in real time through the joint encoder on the upper limb exoskeleton device and the forward and inverse kinematics model of the exoskeleton. The DMP parameters are dynamically adjusted according to the real-time data to ensure the accuracy and smoothness of the motion trajectory. The trained DMP model is used to generate a motion trajectory that conforms to the actual condition of the target patient.

[0111] The process involves setting multiple specific bi-arm collaborative rehabilitation tasks, which are completed using the upper limb exoskeleton device, allowing for collaboration between the healthy and affected sides. Evaluation results are obtained based on the completion rate of each rehabilitation task, specifically including:

[0112] Set up multiple specific bi-arm collaborative rehabilitation tasks and ensure that each rehabilitation task is relevant to the daily life of the target patient;

[0113] The upper limb exoskeleton device is used to complete rehabilitation tasks in collaboration between the healthy and affected sides, determine the completion rate of each rehabilitation task, and obtain evaluation results based on the completion rate of each rehabilitation task.

[0114] The assessment results, obtained based on the completion rate of each rehabilitation task, further include:

[0115] The parameters of the DMP model and the rehabilitation training program are adjusted based on the evaluation results to achieve personalized rehabilitation training.

[0116] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a healthy-affected side collaborative rehabilitation program based on upper limb exoskeleton task-driven method, wherein when the healthy-affected side collaborative rehabilitation program based on upper limb exoskeleton task-driven method is executed by a processor, it implements the steps of the healthy-affected side collaborative rehabilitation method based on upper limb exoskeleton task-driven method as described above.

[0117] In summary, this invention provides a method, system, terminal, and computer-readable storage medium for collaborative rehabilitation of the healthy and affected sides based on task-driven upper limb exoskeleton. The method includes: acquiring patient disease information of a target patient; assessing the upper limb functional status of the target patient based on the patient disease information; matching the required upper limb exoskeleton device according to the characteristics of the target patient and wearing a surface-mounted electromyography (EMG) sensor; collecting surface EMG signals from target muscle groups on the healthy side of the target patient; extracting key features of the surface EMG signals; capturing motion data of the healthy arm dragging the upper limb exoskeleton device; establishing a muscle model; estimating the movement intention based on the key features of the surface EMG signals and the motion data using the muscle model; inputting position sensor data from the upper limb exoskeleton device into a DMP model to complete the affected side imitation learning; and using the trained DMP model to generate a movement trajectory that conforms to the patient's actual condition; setting multiple specific bi-arm collaborative rehabilitation tasks; completing each collaborative rehabilitation task between the healthy and affected sides using the upper limb exoskeleton device; and obtaining an evaluation result based on the completion rate of each rehabilitation task. This invention enables patients to coordinate their healthy and affected sides while completing tasks, promoting the functional recovery of the damaged limbs. Through repeated practice of asymmetrical movements, it enhances the patient's motor control and bilateral limb coordination during later rehabilitation, ultimately restoring the basic motor functions of the damaged limbs. This makes the rehabilitation training process more aligned with the patient's rehabilitation process, improving rehabilitation outcomes while satisfying human-computer interaction requirements.

[0118] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or terminal that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or terminal. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or terminal that includes that element.

[0119] Of course, those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware (such as a processor, controller, etc.). The program can be stored in a computer-readable storage medium, and when executed, it can include the processes described in the above method embodiments. The computer-readable storage medium can be a memory, magnetic disk, optical disk, etc.

[0120] It should be understood that the application of the present invention is not limited to the examples above. Those skilled in the art can make improvements or modifications based on the above description, and all such improvements and modifications should fall within the protection scope of the appended claims.

Claims

1. A collaborative rehabilitation method for the healthy and affected sides based on a task-driven upper limb exoskeleton, characterized in that, The healthy-affected side collaborative rehabilitation method based on upper limb exoskeleton task-driven approach includes: Obtain the patient's disease information, assess the upper limb functional status of the target patient based on the patient's disease information, match the upper limb exoskeleton device to be worn based on the characteristics of the target patient, and wear a surface-mounted electromyography sensor; Surface electromyography (EMG) signals from the target muscle group on the healthy side of the target patient are collected, key features of the EMG signals are extracted, motion data of the healthy arm dragging the upper limb exoskeleton device are captured, a muscle model is established, and motion intention is estimated based on the key features of the EMG signals and the motion data using the muscle model. The position sensor data on the upper limb exoskeleton device is input into the DMP model to complete the imitation learning of the affected side, and the trained DMP model is used to generate a motion trajectory that matches the patient's actual condition. Multiple specific bi-arm collaborative rehabilitation tasks are set, and each rehabilitation task involving the healthy side and the affected side is completed through the upper limb exoskeleton device. Evaluation results are obtained based on the completion rate of each rehabilitation task. The process involves collecting surface electromyography (EMG) signals from the target muscle groups on the healthy side of the target patient, extracting key features from the EMG signals, capturing motion data of the healthy arm dragging the upper limb exoskeleton device, establishing a muscle model, and estimating the movement intention based on the key features of the EMG signals and the motion data using the muscle model. Specifically, this includes: Surface electromyography (EMG) signals from the target muscle group on the healthy side of the target patient are collected. Noise in the surface EMG signals is removed using a digital filter, and key features of the surface EMG signals are extracted and saved. The target patient is instructed to use their unaffected arm to drag the upper limb exoskeleton device and move it, and the motion data during the movement is recorded. The motion data includes the position of the joints, torque, and six-dimensional force values ​​at the upper limb interaction points. In the OpenSim environment, the joint positions, torques, and six-dimensional force values ​​of the upper limb interaction points captured by the motion capture of the dragging upper limb exoskeleton device are imported. The joint torques are derived using the inverse dynamics model of the upper limb exoskeleton, the torques generated by muscle activity are estimated, and the parameters are adjusted to make the model output match the experimental data. Establish a muscle model and optimize the muscle parameters of the muscle model, including optimal fiber length, tendon relaxation length, muscle belly angle, default activation, default fiber length, and contraction speed coefficient. The muscle model estimates the movement intention based on the key features of the surface electromyography signal, and analyzes the specific action that the target patient wants to perform.

2. The method for collaborative rehabilitation of the healthy and affected sides based on upper limb exoskeleton task-driven rehabilitation according to claim 1, characterized in that, The process of acquiring the target patient's disease information, assessing the target patient's upper limb functional status based on the disease information, matching the required upper limb exoskeleton device based on the target patient's characteristics, and wearing a surface-mounted electromyography (EMG) sensor specifically includes: Obtain the target patient's disease information and assess the target patient's upper limb functional status based on the patient's disease information using the Fugl-Meyer scale, a clinical assessment tool. Based on the characteristics of the target patient, a size-matched upper limb exoskeleton device is determined. The upper limb exoskeleton device is used to cover the upper limb joints on the affected and unaffected sides of the target patient, and surface-mounted electromyography (EMG) sensors are worn at designated key muscle locations on the unaffected and affected sides of the target patient.

3. The collaborative rehabilitation method for the healthy and affected sides based on upper limb exoskeleton task-driven rehabilitation according to claim 1, characterized in that, The muscle model includes an SVM classification model; The muscle model estimates motor intent based on key features of the surface electromyography signals, specifically including: During the training phase, an SVM classification model is constructed using known motion actions and corresponding EMG features. During the testing phase, new EMG data is introduced to evaluate the accuracy and generalization ability of the SVM classification model, ensuring that the SVM classification model works correctly on the new data. After model training and validation are completed, the trained SVM classification model is used to predict the target patient's movement intention based on the key features of the surface electromyography signal.

4. The collaborative rehabilitation method for the healthy and affected sides based on upper limb exoskeleton task-driven rehabilitation according to claim 1, characterized in that, The step of inputting position sensor data from the upper limb exoskeleton device into the DMP model to complete the affected side's imitation learning, and using the trained DMP model to generate a motion trajectory that matches the patient's actual condition, specifically includes: The position sensor data from the upper limb exoskeleton device is input into the DMP model, and a Gaussian function is selected as the basis function to determine the number, center position, and width of the basis function. The weights of the basis functions are calculated using the least squares method, so that the generated motion trajectory is consistent with the actual motion trajectory of the target patient. Adjust the natural frequency parameters to control the rhythm of the movement, and adjust the impedance coefficient to control the contact force between the target patient and the exoskeleton. During the DMP model training process, the motion state of the target patient is monitored in real time through the joint encoder on the upper limb exoskeleton device and the forward and inverse kinematics model of the exoskeleton. The DMP parameters are dynamically adjusted according to the real-time data to ensure the accuracy and smoothness of the motion trajectory. The trained DMP model is used to generate a motion trajectory that conforms to the actual condition of the target patient.

5. The collaborative rehabilitation method for the healthy and affected sides based on upper limb exoskeleton task-driven rehabilitation according to claim 4, characterized in that, The system sets up multiple specific bi-arm collaborative rehabilitation tasks, which are completed using the upper limb exoskeleton device. Each task involves collaboration between the healthy and affected sides, and an evaluation result is obtained based on the completion rate of each task. Specifically, this includes: Set up multiple specific bi-arm collaborative rehabilitation tasks and ensure that each rehabilitation task is relevant to the daily life of the target patient; The upper limb exoskeleton device is used to complete rehabilitation tasks in collaboration between the healthy and affected sides, determine the completion rate of each rehabilitation task, and obtain evaluation results based on the completion rate of each rehabilitation task.

6. The collaborative rehabilitation method for the healthy and affected sides based on upper limb exoskeleton task-driven rehabilitation according to claim 1 or 5, characterized in that, The assessment results, obtained based on the completion of each rehabilitation task, also include: The parameters of the DMP model and the rehabilitation training program are adjusted based on the evaluation results to achieve personalized rehabilitation training.

7. A collaborative rehabilitation system for the healthy and affected sides based on a task-driven upper limb exoskeleton, characterized in that, The task-driven, healthy-affected-side collaborative rehabilitation system based on an upper limb exoskeleton is used to implement the task-driven, healthy-affected-side collaborative rehabilitation method based on an upper limb exoskeleton as described in any one of claims 1-6. The task-driven, healthy-affected-side collaborative rehabilitation system based on an upper limb exoskeleton comprises: The assessment and matching module is used to acquire the patient's disease information, assess the upper limb functional status of the target patient based on the patient's disease information, match the upper limb exoskeleton device to be worn based on the characteristics of the target patient, and wear a surface-mount electromyography sensor. The acquisition and estimation module is used to acquire surface electromyography (EMG) signals from the target muscle group on the healthy side of the target patient, extract key features of the surface EMG signals, capture motion data of the healthy arm dragging the upper limb exoskeleton device, establish a muscle model, and estimate the motion intention based on the key features of the surface EMG signals and the motion data through the muscle model. The learning and generation module is used to input the position sensor data of the upper limb exoskeleton device into the DMP model to complete the imitation learning of the affected side, and use the trained DMP model to generate a motion trajectory that conforms to the patient's actual condition. The rehabilitation and assessment module is used to set multiple specific bi-arm collaborative rehabilitation tasks. Each rehabilitation task involving the healthy and affected sides is completed through the upper limb exoskeleton device, and assessment results are obtained based on the completion rate of each rehabilitation task.

8. A terminal, characterized in that, The terminal includes: a memory, a processor, and a task-driven healthy-affected side collaborative rehabilitation program based on an upper limb exoskeleton, stored in the memory and executable on the processor. When the task-driven healthy-affected side collaborative rehabilitation program based on an upper limb exoskeleton is executed by the processor, it implements the steps of the task-driven healthy-affected side collaborative rehabilitation method based on an upper limb exoskeleton as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a healthy-affected side collaborative rehabilitation program based on upper limb exoskeleton task-driven method, which, when executed by a processor, implements the steps of the healthy-affected side collaborative rehabilitation method based on upper limb exoskeleton task-driven method as described in any one of claims 1-6.

Images