Auxiliary rehabilitation method based on multi-source physiological signal wearable device and related device

Abstract

The invention discloses an auxiliary rehabilitation method based on a multi-source physiological signal wearable device and a related device.According to the method, an arm band, electric robot gloves and an ankle rehabilitation instrument are included, and the arm band is used for collecting forearm neuromuscular signals; the electric robot glove is used for collecting hand electrophysiological signals and assisting hand movement, the ankle rehabilitation instrument is used for collecting ankle electric signals, balancing sole pressure and forcibly pedaling and stepping, and the ankle rehabilitation instrument comprises a pressure sensor and neuromuscular electrical stimulation equipment; inputting the forearm neuromuscular signal and the hand electrophysiological signal into an arm posture and gait model, and predicting an arm movement posture and a foot movement state of the patient; the electric robot glove synchronously assists hand movement according to the predicted arm movement posture; and the ankle rehabilitation instrument synchronously assists the foot movement and corrects the ankle according to the predicted foot movement state. According to the invention, the multi-source physiological signals are fused through the arm posture and gait model, and an accurate and personalized auxiliary rehabilitation method is provided.

Description

Technical Field

[0001] The present invention relates to the field, and specifically relates to an assisted rehabilitation method, device, and computing device based on a multi-source physiological signal wearable device. Background Art

[0002] Stroke patients often face problems such as balance dysfunction, muscle strength decline, and sensory abnormalities, and are more prone to falls. The rehabilitation goal after stroke is to restore the patient's functions (including motor, sensory, cognitive, language, and social participation abilities) to the greatest extent.

[0003] Existing stroke rehabilitation methods include physical therapy, occupational therapy, etc. The rehabilitation effect of patients is affected by various factors (such as stroke severity, lesion location, age, basic health status, etc.), often relying on the guidance of therapists. The patient's participation degree is not high, and it is usually relatively subjective, making it difficult to accurately quantify the patient's functional impairment and rehabilitation progress, and difficult to predict and control.

[0004] To solve the above problems, the present invention proposes an assisted rehabilitation method based on a multi-source physiological signal wearable device, which fuses multi-source physiological signals through an arm posture and gait model to provide a precise and personalized assisted rehabilitation method. Summary of the Invention

[0005] In view of the above problems, the present invention provides an assisted rehabilitation method, device, and computing device based on a multi-source physiological signal wearable device.

[0006] According to one aspect of the present invention, there is provided an assisted rehabilitation method based on a multi-source physiological signal wearable device, including:

[0007] An armband, an electric robot glove, and an ankle-foot rehabilitation instrument, wherein the armband is used to collect forearm neuromuscular signals; the electric robot glove is used to collect hand electrophysiological signals and assist hand movements, and the ankle-foot rehabilitation instrument is used to collect ankle-foot electrical signals and balance the sole pressure and forcefully push off the ground to take a step. The ankle-foot rehabilitation instrument includes a plurality of pixel pressure sensors and neuromuscular electrical stimulation devices;

[0008] Input the forearm neuromuscular signals and the hand electrophysiological signals into a trained arm posture and gait model to predict the patient's arm movement posture and foot movement state;

[0009] The electric robot glove synchronously assists hand movements according to the predicted arm movement posture; the ankle-foot rehabilitation instrument synchronously assists foot movements and corrects the ankle according to the predicted foot movement state.

[0010] In an alternative embodiment, the forearm neuromuscular signals include: electromyogram signals reflecting the contraction state of muscle fibers and muscle tension signals reflecting the muscle tension and relaxation states;

[0011] The hand electrophysiological signals include: hand electromyogram signals reflecting the contraction and relaxation states of hand muscles and skin conductance signals caused by changes in skin sweat secretion;

[0012] The ankle electrical signals include: ankle electromyogram signals reflecting the contraction and relaxation states of ankle muscles and plantar pressure signals of the foot reflecting the force state between the sole and the ground.

[0013] In an alternative embodiment, the arm posture and gait model includes:

[0014] An input layer for inputting input vectors of forearm neuromuscular signals, hand electrophysiological signals, and ankle electrical signals;

[0015] Two layers of CNN network layers, where the first layer of CNN network layer is used to extract local features of the input vector, and the second layer of CNN network layer is used to extract global features of the input vector;

[0016] A Depth CNN network layer for extracting spatio-temporal features of local and global features;

[0017] A Separable CNN network layer for convolving and fusing spatio-temporal features of each input channel;

[0018] A Bi-LSTM network layer for capturing the temporal dependence relationships of spatio-temporal features;

[0019] A DCA network layer for calculating the covariance matrix of the temporal dependence relationships and decomposing it;

[0020] An output layer for predicting the arm movement posture and foot movement state of the patient.

[0021] In an alternative embodiment, the electric robot glove further synchronously assisting hand movement according to the predicted arm movement posture includes:

[0022] The electric robot glove drives the joints of the fingers and wrist to move through a motor device according to the predicted arm movement posture, so that they are consistent with the predicted arm movement posture.

[0023] In an alternative embodiment, the ankle rehabilitation device further synchronously assisting foot movement and correcting the ankle according to the predicted foot movement state includes:

[0024] The ankle rehabilitation device adjusts the angle, applied force, and movement speed of the rehabilitation device through an actuator according to the predicted foot movement state and the patient's physical characteristics; wherein, the patient's physical characteristics include age, gender, height, weight, and rehabilitation status.

[0025] In an alternative manner, the step of making it consistent with the predicted arm movement posture further includes:

[0026] The electric robot glove establishes the arm kinematic equations of the fingers and wrist joints through the Denavit-Hartenberg parameter method;

[0027] Solve the arm kinematic equations according to the predicted arm movement posture to obtain the target angles or positions of each finger and wrist joint;

[0028] Generate the movement trajectories of the joints according to the target angles or positions of each finger and wrist joint through the B-spline curve algorithm method;

[0029] According to the implementation of the joint movement trajectories, automatically adjust the control parameters of the electric robot glove through an adaptive fuzzy control algorithm to be consistent with the predicted arm movement posture.

[0030] In an alternative manner, the step that the ankle rehabilitation device adjusts the angle, applied force, and movement speed of the rehabilitation device through an actuator according to the predicted foot movement state and the patient's physical characteristics further includes:

[0031] The ankle rehabilitation device inputs the predicted foot movement state and the patient's physical characteristics into the ankle movement equation;

[0032] The ankle movement equation calculates the joint angles, speeds, and accelerations according to the inverse dynamics principle.

[0033] In an alternative manner, the expression of the arm kinematic equation is:

[0034]

[0035] where τ is the torque vector of the ankle joint, representing the force or torque required to generate joint movement; M(q) is the inertia matrix, which is a function of the ankle inertia characteristics with respect to the joint angle q; is the joint acceleration vector, representing the rate of change of the joint angle with time; is the centrifugal force and Coriolis force vector, which is the additional force generated by the rotational movement of the ankle; G(q) is the gravity vector; is a function of the joint angle, speed, and patient physical characteristic parameters;

[0036] The expression of the ankle movement equation is:

[0037] T j = A 1 A 2 …A j

[0038] Wherein, T j is the transformation matrix from the base to the i-th joint; A j is the DH transformation matrix of the j-th joint; θ j is the rotation angle around z j-1 ; α j is the deflection angle from the z j-1 axis to the z j axis; d j is the distance from the x j-1 axis to the x j axis along the z j-1 axis.

[0039] According to another aspect of the present invention, there is provided an assisted rehabilitation device based on a multi-source physiological signal wearable device, including:

[0040] A data acquisition module, including an armband, an electric robot glove and an ankle-foot rehabilitation instrument. Among them, the armband is used to collect forearm neuromuscular signals; the electric robot glove is used to collect hand electrophysiological signals and assist hand movements, and the ankle-foot rehabilitation instrument is used to collect ankle-foot electrical signals and balance the sole pressure and forceful stepping on the ground. The ankle-foot rehabilitation instrument includes a plurality of pixel pressure sensors and neuromuscular electrical stimulation devices;

[0041] An attitude prediction module, configured to input the forearm neuromuscular signals and the hand electrophysiological signals into a trained arm posture and gait model to predict the patient's arm movement posture and foot movement state;

[0042] An assisted rehabilitation module, configured to the electric robot glove synchronously assist hand movements according to the predicted arm movement posture; the ankle-foot rehabilitation instrument synchronously assists foot movements and corrects the ankle according to the predicted foot movement state.

[0043] According to still another aspect of the present invention, there is provided a computing device, including: a processor, a memory, a communication interface and a communication bus, and the processor, the memory and the communication interface complete communication with each other through the communication bus;

[0044] The memory is used to store at least one executable instruction, and the executable instruction causes the processor to execute the operations corresponding to the above-mentioned assisted rehabilitation method based on a multi-source physiological signal wearable device.

[0045] According to the solution provided by the present invention, it includes an armband, an electric robotic glove, and an ankle rehabilitation device. Among them, the armband is used to collect forearm neuromuscular signals; the electric robotic glove is used to collect hand electrophysiological signals and assist hand movement, and the ankle rehabilitation device is used to collect ankle electrical signals and balance the sole pressure and force to step forward by pushing the ground. The ankle rehabilitation device includes a plurality of pixel pressure sensors and a neuromuscular electrical stimulation device; the forearm neuromuscular signals and the hand electrophysiological signals are input into a trained arm posture and gait model to predict the patient's arm movement posture and foot movement state; the electric robotic glove synchronously assists hand movement according to the predicted arm movement posture; the ankle rehabilitation device synchronously assists foot movement and corrects the ankle according to the predicted foot movement state. The present invention fuses multi-source physiological signals through an arm posture and gait model to provide a precise and personalized assisted rehabilitation method.

[0046] The above description is only an overview of the technical solution of the present invention. In order to be able to understand the technical means of the present invention more clearly, it can be implemented according to the content of the specification. And in order to make the above and other purposes, features and advantages of the present invention more obvious and understandable, the following specifically describes the embodiments of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered to be a limitation of the present invention. And throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0048] Figure 1 A flowchart showing the assisted rehabilitation method of the wearable device based on multi-source physiological signals according to an embodiment of the present invention;

[0049] Figure 2 A schematic diagram showing the armband and the electric robotic glove according to an embodiment of the present invention;

[0050] Figure 3 A schematic diagram showing the framework of the assisted rehabilitation according to an embodiment of the present invention;

[0051] Figure 4 A schematic diagram showing the arm posture and gait model according to an embodiment of the present invention;

[0052] Figure 5 A schematic diagram showing the framework of the assisted rehabilitation device based on multi-source physiological signals according to an embodiment of the present invention;

[0053] Figure 6 A schematic diagram showing the structure of the computing device according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0054] Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present invention can be more thoroughly understood and the scope of the present invention can be fully conveyed to those skilled in the art.

[0055] Figure 1 The flowchart of the assisted rehabilitation method of the multi-source physiological signal wearable device according to the embodiment of the present invention is shown. Specifically, as Figure 1 shown, it includes the following steps:

[0056] Step S101, collecting forearm neuromuscular signals through an armband; collecting hand electrophysiological signals through an electric robotic glove and assisting hand movement, and collecting ankle electrical signals through an ankle rehabilitation device and balancing the sole pressure and pushing force to step forward. The above-mentioned ankle rehabilitation device includes a plurality of pixel pressure sensors and neuromuscular electrical stimulation devices.

[0057] In this embodiment, the framework schematic diagram of the assisted rehabilitation method based on the multi-source physiological signal wearable device is as Figure 3 shown. Using sensors on the wearable device, such as forearm muscle signal sensors, gait monitoring sensors, etc., to collect muscle tension signals, electromyogram signals, and posture information of the patient. According to the collected data, an arm posture model and a gait model are constructed. Among them, the arm posture model is used to analyze the movement state of the arm, including muscle activity, joint angle, etc., and the gait model is used to analyze the walking posture and gait characteristics of the patient. Extract features from the collected data and construct an arm kinematic equation and an ankle kinematic equation, and use a controller to generate a control signal to assist the hand movement and foot movement of the patient through a wearable assistive device.

[0058] Specifically, as Figure 2 shown, through an armband, an electric robotic glove, and an ankle rehabilitation device, forearm neuromuscular signals, hand electrophysiological signals, and ankle electrical signals are collected simultaneously, which can more comprehensively and accurately reflect the movement state and nerve function of the patient, so as to achieve more precise rehabilitation assistance. Among them, the armband is worn on the forearm, and the neuromuscular signals of the forearm muscles are collected through a surface electromyogram (sEMG) sensor to reflect the movement intention and muscle activity state of the patient. The electric robotic glove is worn on the hand, and the muscle and movement data of the hand are collected through electrophysiological sensors (such as electromyogram, capacitive sensors, etc.). The ankle rehabilitation device is worn on the foot and ankle, including a plurality of pixel pressure sensors and neuromuscular electrical stimulation devices, and the pressure sensors monitor the sole pressure distribution.

[0059] In an alternative manner, the forearm neuromuscular signals include: an electromyogram signal reflecting the contraction state of muscle fibers and a muscle tension signal reflecting the muscle tension and relaxation states;

[0060] The hand electrophysiological signals include: a hand electromyogram signal reflecting the contraction and relaxation states of hand muscles and a skin conductance signal representing the change in conductivity caused by skin sweat secretion;

[0061] The ankle-foot electrical signals include: an ankle-foot electromyogram signal reflecting the contraction and relaxation states of ankle-foot muscles and a plantar pressure signal of the foot reflecting the force state between the sole of the foot and the ground.

[0062] In this embodiment, specifically, the electromyogram signal collects the fiber contraction state of the forearm muscles through a surface electromyogram (sEMG) sensor, the muscle tension signal collects the signal of the muscle tension and relaxation states through a tension sensor (such as a piezoelectric sensor), the hand electrophysiological signal collects the contraction and relaxation states of the hand muscles through an sEMG sensor worn on the hand, the skin conductance signal collects the change in conductivity caused by skin sweat secretion through a skin conductance sensor (such as a sweat conductance sensor), the ankle-foot electromyogram signal collects the contraction and relaxation states of the ankle-foot muscles through an sEMG sensor worn on the ankle-foot, and the plantar pressure signal monitors the force state between the sole of the foot and the ground through a plurality of pixel pressure sensors.

[0063] Step S102: Input the forearm neuromuscular signals and the hand electrophysiological signals into a trained arm posture and gait model to predict the arm movement posture and foot movement state of the patient.

[0064] In an alternative manner, the arm posture and gait model includes:

[0065] An input layer for inputting input vectors of forearm neuromuscular signals, hand electrophysiological signals, and ankle-foot electrical signals;

[0066] Two layers of CNN network layers, where the first layer of CNN network layer is used to extract local features of the input vector, and the second layer of CNN network layer is used to extract global features of the input vector;

[0067] A Depth CNN network layer for extracting spatio-temporal features of local features and global features;

[0068] A Separable CNN network layer for convolving and fusing the spatio-temporal features of each input channel;

[0069] A Bi-LSTM network layer for capturing the temporal dependence relationships of spatio-temporal features;

[0070] A DCA network layer for calculating the covariance matrix of the temporal dependence relationships and decomposing it;

[0071] An output layer for predicting the arm movement posture and foot movement state of a patient.

[0072] In this embodiment, as Figure 4 shown, a multi-layer CNN network layer is used to extract local and global features of the input signal, and a Depth CNN network layer is further used to extract spatio-temporal features. The Bi-LSTM network layer captures temporal dependencies, enabling a better understanding of the temporal variation law of dynamic signals. The Separable CNN network layer and the DCA network layer perform convolution and fusion on the spatio-temporal features of each input channel, calculate the covariance matrix of the temporal dependencies and decompose it, thereby improving the overall performance of the model.

[0073] Step S103, the electric robot glove synchronously assists hand movement according to the predicted arm movement posture; the ankle-foot rehabilitation device synchronously assists foot movement and corrects the ankle according to the predicted foot movement state.

[0074] In this embodiment, the electric robot glove synchronously assists hand movement according to the predicted arm movement posture to respond to the user's movement intention, provide necessary assistance or resistance, and help complete specific actions. The ankle-foot rehabilitation device synchronously assists foot movement and corrects the ankle according to the predicted foot movement state, recognizes whether the user is standing, walking or in other movement states, and provides corresponding support, assistance or correction. By synchronously assisting hand and ankle movement, the patient can receive assistance while actively attempting to move, which can better stimulate neuroplasticity, reduce movement compensation, and achieve a more natural and smooth rehabilitation training. It can be applied to the rehabilitation of limb function disorders caused by various neurological diseases such as stroke, cerebral palsy, and spinal cord injury. For example, if the model predicts that the patient wants to bend the elbow joint, the electric glove provides assistance to help the patient complete the bending action. If the patient has abnormal tremors during the bending of the elbow joint, the glove provides resistance to help with smooth movement. If the model predicts that the patient is in a walking state, the ankle-foot rehabilitation device provides foot support and assistance when taking a step. If the patient's ankle has varus or valgus during walking, the rehabilitation device corrects it to help the patient maintain a correct gait.

[0075] In an alternative manner, the electric robot glove synchronously assisting hand movement according to the predicted arm movement posture further includes:

[0076] The electric robot glove drives the joints of the fingers and wrist to move according to the predicted arm movement posture through a motor device, so that they are consistent with the predicted arm movement posture.

[0077] In this embodiment, the arm postures obtained through model prediction include, for example, the elbow joint bending angle, the forearm rotation angle, etc. The electric robot glove uses a motor device to accurately drive the joint movements of parts such as fingers and wrists. The joint movements of the glove are consistent with the predicted arm movement postures, realizing the coordination between hand movements and arm postures. The multi-joint movement can drive the joints of multiple fingers and wrists to realize complex hand movements, such as grasping, pinching, rotating, etc.

[0078] Specifically, the motor drive system designs an independent motor drive system for each joint of the fingers and wrists. Each motor is equipped with an independent drive circuit and a control unit. Force feedback sensors are installed at key parts of the fingers and wrists to monitor the force exerted by the patient in real time. According to the predicted arm postures of the model, the movement trajectories of each joint are calculated through a control algorithm, and the joint movements and force feedback data of the glove are monitored in real time. The output of the motor is adjusted according to the real-time feedback to ensure that the movement trajectory of the glove is consistent with the predicted arm postures.

[0079] In an alternative way, making it consistent with the predicted arm movement postures further includes: the electric robot glove establishes the arm kinematic equations of the finger and wrist joints through the Denavit-Hartenberg parameter method;

[0080] Solving the arm kinematic equations according to the predicted arm movement postures to obtain the target angles or positions of each finger and wrist joint;

[0081] Generating the movement trajectories of the joints according to the target angles or positions of each finger and wrist joint through the B-spline curve algorithm method;

[0082] According to the implementation of the joint movement trajectories, automatically adjust the control parameters of the electric robot glove through an adaptive fuzzy control algorithm to be consistent with the predicted arm movement postures.

[0083] In this embodiment, the Denavit-Hartenberg (D-H) parameter method establishes the arm kinematic equations through D-H parameters, which can accurately describe the movement structures of the arm and fingers. By solving the kinematic equations, according to the desired arm movement postures, the target angles or positions of each joint are accurately calculated. The B-spline curve generates smooth joint movement trajectories, avoiding sudden jumps and jerky movements, ensuring the smoothness and coordination of the movement. The adaptive fuzzy control dynamically adjusts the control parameters according to the actual movement situation, overcomes the limitation of fixed parameters in traditional control methods, and improves the adaptability of the control system. Through the smooth movement trajectories and adaptive control parameters, the movement of the electric robot glove is closer to the natural movement of the human body, improving the comfort of the wearer and the intuitiveness of the control.

[0084] Specifically, according to the structure of the electric robot glove, the connection relationship and geometric parameters between each joint are determined, and a DH parameter table is established. Based on the DH parameters, a kinematic equation from the fingertips to the wrist is established to describe the motion relationship of each joint.

[0085] Based on the collected data of external sensors (such as inertial measurement unit IMU, electromyography sensor EMG, etc.), the user's desired arm movement posture (for example, reaching a target position or performing a specific gesture) is predicted.

[0086] According to the predicted arm movement posture, the kinematic equations are solved using an inverse kinematics algorithm (eg, a numerical iterative method or an analytical solution) to obtain the target angles or positions of the various fingers and wrist joints.

[0087] Using the B-spline curve algorithm, the initial angle / position and target angle / position of the joint are used as control points to generate a smooth joint motion trajectory.

[0088] An adaptive fuzzy control algorithm is used to dynamically adjust the control parameters (e.g., the driving force / torque of the motor) of the electric robot glove according to the deviation between the actual motion state of the joint (e.g., angle / position, speed, acceleration, etc.) and the motion trajectory. Among them, the fuzzy rules predefine fuzzy rules to describe the relationship between the deviation and the control parameters.

[0089] Through online learning and adjustment, the fuzzy rules and control parameters are continuously optimized to achieve precise control of the electric robot glove and keep its motion trajectory consistent with the predicted arm motion posture as much as possible. The adaptive fuzzy controller generates control signals to drive the motor of the electric robot glove and control the movement of each joint.

[0090] In an optional manner, the foot and ankle rehabilitation device synchronously assists foot movement and corrects the foot and ankle according to the predicted foot movement state, further comprising:

[0091] The foot and ankle rehabilitation device adjusts the angle, applied force and movement speed of the rehabilitation device through an actuator according to the predicted foot movement state and the patient's physical characteristics; wherein the patient's physical characteristics include age, gender, height, weight and rehabilitation state.

[0092] In this embodiment, a more personalized rehabilitation plan can be provided by considering the patient's age, gender, height, weight and rehabilitation status. According to the predicted foot movement state, the actuator dynamically adjusts the angle, applied force and movement speed of the rehabilitation device to provide more effective foot assistance and correction.

[0093] In an alternative manner, the ankle rehabilitation device further includes adjusting the angle, applied force, and movement speed of the rehabilitation device through an actuator according to the predicted foot movement state and the physical characteristics of the patient: the ankle rehabilitation device inputs the predicted foot movement state and the physical characteristics of the patient into the ankle movement equation; the ankle movement equation calculates the joint angles, speeds, and accelerations based on the inverse dynamics principle. In this embodiment, the predicted foot movement state and the physical characteristics of the patient are input into the ankle movement equation, and the joint angles, speeds, and accelerations are calculated based on the inverse dynamics principle. Specifically, the Denavit-Hartenberg (D-H) parameter method is used to establish the dynamic equation of ankle movement. The angle of the rehabilitation device is adjusted according to the calculated joint angle to make it consistent with the expected movement posture, avoiding too fast or too slow movement.

[0094] In an alternative manner, the expression of the arm kinematic equation is:

[0095]

[0096] where τ is the torque vector of the ankle joint, representing the force or torque required to generate joint movement; M(q) is the inertia matrix, which is a function of the ankle inertia characteristics with respect to the joint angle q; is the joint acceleration vector, representing the rate of change of the joint angle with respect to time; is the centrifugal force and Coriolis force vector, which is the additional force generated by the rotational movement of the ankle; G(q) is the gravity vector; is a function of the joint angle, speed, and physical characteristic parameters of the patient;

[0097] The expression of the ankle movement equation is:

[0098] T j =A 1 A 2 …A j

[0099] where T j is the transformation matrix from the base to the i-th joint; A j is the DH transformation matrix of the j-th joint; θ j is the rotation angle about z j-1 ; α j is the deflection angle from the z j-1 axis to the z j axis; d j is the distance along the z j-1 axis from the x j axis to the x j-1 axis.

[0100] In this embodiment, the dynamic equation takes into account inertial force, centrifugal force / Coriolis force, gravity, and external force, and can more accurately describe the motion state of the ankle joint. The item includes the patient's physical characteristic parameters, which are adjusted according to the individual differences of the patients to achieve personalized rehabilitation. The D-H parameter method is used to establish the kinematic equation to describe the relative positions and postures between each joint. The transformation matrix is used to describe the motion of each joint, making the kinematic equation clearer.

[0101] Specifically, the inertial parameters of the ankle, such as mass, moment of inertia, etc., are determined through inverse dynamics experiments. The centrifugal force and Coriolis force are calculated through kinematic relationships and angular velocity and angular acceleration. The gravity vector (such as the gravity moment) is calculated based on the mass of the ankle, the position of the center of gravity, and the acceleration due to gravity.

[0102] Data such as joint angle, velocity, acceleration, and force are collected in real time, and the collected data is input into the control algorithm to calculate the required control torque. The motor drive mechanism is used to generate the calculated control torque. According to the structure of the ankle joint, the coordinate system of each joint is defined. According to the defined coordinate system, the DH parameters of each joint are determined, and the transformation matrix of each joint is calculated according to the DH parameters.

[0103] The transformation matrices of all joints are multiplied in sequence to obtain the transformation matrix from the base to the end of the ankle. The position coordinates (x, y, z) and attitude information (such as using a rotation matrix or Euler angles to represent the attitude) of the end of the ankle are extracted from the transformation matrix. The read angle values are substituted into the equation for calculation, so as to obtain the position and attitude of the end of the ankle in real time.

[0104] According to the solution provided by the present invention, it includes an armband, an electric robot glove, and an ankle rehabilitation device. Among them, the armband is used to collect forearm neuromuscular signals; the electric robot glove is used to collect hand electrophysiological signals and assist hand movement, and the ankle rehabilitation device is used to collect ankle electrical signals and balance the sole pressure and force to step on the ground. The ankle rehabilitation device includes a plurality of pixel pressure sensors and neuromuscular electrical stimulation devices; the forearm neuromuscular signals and the hand electrophysiological signals are input into the trained arm posture and gait model to predict the patient's arm movement posture and foot movement state; the electric robot glove synchronously assists hand movement according to the predicted arm movement posture; the ankle rehabilitation device synchronously assists foot movement and corrects the ankle according to the predicted foot movement state. The present invention fuses multi-source physiological signals through the arm posture and gait model to provide a precise and personalized assisted rehabilitation method. Figure 5 The frame schematic diagram of the assisted rehabilitation device based on the multi-source physiological signal wearable device according to the embodiment of the present invention is shown. The assisted rehabilitation device based on the multi-source physiological signal wearable device includes:

[0105] The data acquisition module 510 includes an armband, an electric robotic glove, and an ankle rehabilitation device. Among them, the armband is used to collect forearm neuromuscular signals; the electric robotic glove is used to collect hand electrophysiological signals and assist hand movement, and the ankle rehabilitation device is used to collect ankle electrical signals and balance the sole pressure and force to step on the ground. The ankle rehabilitation device includes a plurality of pixel pressure sensors and neuromuscular electrical stimulation devices;

[0106] The posture prediction module 520 is used to input the forearm neuromuscular signals and the hand electrophysiological signals into the trained arm posture and gait model to predict the patient's arm movement posture and foot movement state;

[0107] The assisted rehabilitation module 530 is used for the electric robotic glove to assist hand movement synchronously according to the predicted arm movement posture; the ankle rehabilitation device assists foot movement and corrects the ankle synchronously according to the predicted foot movement state.

[0108] Figure 6 The structure diagram of the computing device embodiment of the present invention is shown. The specific implementation of the present invention does not limit the specific implementation of the computing device.

[0109] As Figure 6 shown, the computing device may include: a processor 602, a communication interface 604, a memory 606, and a communication bus 608. Among them: the processor 602, the communication interface 604, and the memory 606 communicate with each other through the communication bus 608. The communication interface 604 is used to communicate with network elements of other devices such as clients or other servers. The processor 602 is used to execute the program 610, and specifically can execute the relevant steps in the above-mentioned assisted rehabilitation method embodiment based on the multi-source physiological signal wearable device.

[0110] Specifically, the program 610 may include program code, and the program code includes computer operation instructions. The processor 602 may be a central processing unit CPU, or a specific integrated circuit ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present invention. One or more processors included in the computing device may be of the same type of processor, such as one or more CPUs; or may be different types of processors, such as one or more CPUs and one or more ASICs.

[0111] A memory 606 for storing a program 610. The memory 606 may include high-speed RAM memory and may also include non-volatile memory, such as at least one disk memory.

[0112] The solution provided by the present invention includes an armband, an electric robot glove, and an ankle rehabilitation device. Among them, the armband is used to collect forearm neuromuscular signals; the electric robot glove is used to collect hand electrophysiological signals and assist hand movement, and the ankle rehabilitation device is used to collect ankle electrical signals and balance the sole pressure and forcefully push off the ground to take a step. The ankle rehabilitation device includes a plurality of pixel pressure sensors and neuromuscular electrical stimulation devices; the forearm neuromuscular signals and the hand electrophysiological signals are input into a trained arm posture and gait model to predict the patient's arm movement posture and foot movement state; the electric robot glove synchronously assists hand movement according to the predicted arm movement posture; the ankle rehabilitation device synchronously assists foot movement and corrects the ankle according to the predicted foot movement state. The present invention provides a precise and personalized assisted rehabilitation method by fusing multi-source physiological signals through an arm posture and gait model. Those skilled in the art can understand that the modules in the devices in the embodiments can be adaptively changed and arranged in one or more devices different from this embodiment. The modules or units or components in the embodiments can be combined into one module or unit or component, and in addition, they can be divided into multiple sub-modules or sub-units or sub-components. Except that at least some of such features and / or processes or units are mutually exclusive, any combination can be used to combine all the features disclosed in this specification (including the accompanying claims, abstract, and drawings) and all the processes or units of any method or device so disclosed. Unless otherwise explicitly stated, each feature disclosed in this specification (including the accompanying claims, abstract, and drawings) can be replaced by an alternative feature that provides the same, equivalent, or similar purpose. In addition, those skilled in the art can understand that although some of the embodiments herein include certain features included in other embodiments but not other features, the combination of the features of different embodiments means that it is within the scope of the present invention and forms different embodiments. For example, in the following claims, any one of the claimed embodiments can be used in any combination. The present invention can be implemented by means of hardware including several different elements and by means of a properly programmed computer. In the unit claims listing several devices, several of these devices can be specifically embodied by the same hardware item. The steps in the above embodiments, unless otherwise specified, should not be construed as a limitation on the execution order.

Claims

1. An assisted rehabilitation method based on a multi-source physiological signal wearable device, characterized in that: include: An armband, an electric robot glove and an ankle rehabilitation device, wherein the armband is used to collect forearm neuromuscular signals; the electric robot glove is used to collect hand electrophysiological signals and assist hand movements; the ankle rehabilitation device is used to collect ankle electrical signals and balance sole pressure and push off the ground to take steps; the ankle rehabilitation device includes multiple pixel pressure sensors and a neuromuscular electrical stimulation device; Inputting the forearm neuromuscular signal and the hand electrophysiological signal into a trained arm posture and gait model to predict the patient's arm movement posture and foot movement state; The electric robot glove synchronously assists hand movement according to the predicted arm movement posture; the foot and ankle rehabilitation device synchronously assists foot movement and corrects the foot and ankle according to the predicted foot movement state.

2. The assisted rehabilitation method based on a multi-source physiological signal wearable device according to claim 1, characterized in that: The forearm neuromuscular signals include: electromyographic signals reflecting the contraction state of muscle fibers and muscle tension signals reflecting the tension and relaxation state of muscles; The hand electrophysiological signals include: hand electromyography signals reflecting the contraction and relaxation states of hand muscles and skin conductance signals of conductivity changes caused by skin sweat secretion; The ankle electrical signals include: ankle electromyography signals reflecting the contraction and relaxation states of the ankle muscles and plantar pressure signals reflecting the stress states of the sole of the foot and the ground.

3. The assisted rehabilitation method based on a multi-source physiological signal wearable device according to claim 2, characterized in that: The arm posture and gait model includes: An input layer, for inputting input vectors of forearm neuromuscular signals, hand electrophysiological signals, and ankle electrical signals; Two layers of CNN network layers, where the first CNN network layer is used to extract local features of the input vector, and the second CNN network layer is used to extract global features of the input vector; Depth CNN network layer, used to extract spatiotemporal features of local features and global features; Separable CNN network layer, used to convolve and fuse the spatiotemporal features of each input channel; Bi-LSTM network layer, used to capture the temporal dependencies of spatiotemporal features; The DCA network layer is used to calculate the covariance matrix of temporal dependencies and decompose them; The output layer is used to predict the patient's arm movement posture and foot movement status.

4. The assisted rehabilitation method based on a multi-source physiological signal wearable device according to claim 1, characterized in that: The electric robot glove synchronously assists hand movement according to the predicted arm movement posture further includes: The electric robot glove drives the joint movements of the fingers and wrist through a motor device according to the predicted arm movement posture, so that the joints are consistent with the predicted arm movement posture.

5. The assisted rehabilitation method based on multi-source physiological signal wearable device according to claim 1, characterized in that: The foot and ankle rehabilitation device synchronously assists foot movement and corrects the foot and ankle according to the predicted foot movement state, further comprising: The foot and ankle rehabilitation device adjusts the angle, applied force and movement speed of the rehabilitation device through an actuator according to the predicted foot movement state and the patient's physical characteristics; wherein the patient's physical characteristics include age, gender, height, weight and rehabilitation state.

6. The assisted rehabilitation method based on a multi-source physiological signal wearable device according to claim 4, characterized in that: The making it consistent with the predicted arm movement posture further includes: The electric robot glove establishes arm kinematic equations of finger and wrist joints through the Denavit-Hartenberg parameter method; Solving the arm kinematics equation according to the predicted arm motion posture to obtain the target angle or position of each finger and wrist joint; The B-spline curve algorithm method is used to generate the motion trajectory of the joints according to the target angle or position of each finger and wrist joint; According to the motion trajectory of the implemented joint, the control parameters of the electric robot glove are automatically adjusted through an adaptive fuzzy control algorithm to be consistent with the predicted arm motion posture.

7. The assisted rehabilitation method based on a multi-source physiological signal wearable device according to claim 5, characterized in that: The foot and ankle rehabilitation device further includes: adjusting the angle, applied force and movement speed of the rehabilitation device through the actuator according to the predicted foot movement state and the patient's physical characteristics: The foot and ankle rehabilitation instrument inputs the predicted foot movement state and the patient's physical characteristics into the foot and ankle movement equation; The ankle motion equation is calculated according to the inverse dynamics principle to obtain the joint angle, velocity and acceleration.

8. The assisted rehabilitation method based on a multi-source physiological signal wearable device according to claim 6 or 7, characterized in that: The expression of the arm kinematic equation is: Where τ is the moment vector of the ankle joint, which represents the force or torque required to produce joint movement; M(q) is the inertia matrix, which is the ankle inertia characteristic function with respect to the joint angle q; is the joint acceleration vector, which represents the rate of change of the joint angle over time; is the centrifugal force and Coriolis force vector, which is the additional force generated by the ankle rotation; G(q) is the gravity vector; It is a function of joint angle, velocity and patient's physical characteristic parameters; The expression of the ankle motion equation is: T j =A1A2…A j Among them, T j is the transformation matrix from the base to the i-th joint; A j DH transformation matrix of the jth joint; θ j It is around z j-1 The rotation angle of j It is from z j-1 Axis to z j The deflection angle of the axis; d j From x j-1 to x j Axis along z j-1 Axis distance.

9. An auxiliary rehabilitation device based on a multi-source physiological signal wearable device, characterized in that: include: The data acquisition module includes an armband, an electric robot glove and an ankle rehabilitation device, wherein the armband is used to collect forearm neuromuscular signals; the electric robot glove is used to collect hand electrophysiological signals and assist hand movements; the ankle rehabilitation device is used to collect ankle electrical signals and balance foot sole pressure and push the ground with force to step, and the ankle rehabilitation device includes multiple pixel pressure sensors and a neuromuscular electrical stimulation device; A posture prediction module, used for inputting the forearm neuromuscular signal and the hand electrophysiological signal into a trained arm posture and gait model to predict the patient's arm movement posture and foot movement state; The auxiliary rehabilitation module is used for the electric robot glove to synchronously assist hand movements according to the predicted arm movement posture; the foot and ankle rehabilitation device synchronously assists foot movements and corrects the foot and ankle according to the predicted foot movement state.

10. A computing device comprising: A processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the above-mentioned auxiliary rehabilitation method based on a multi-source physiological signal wearable device.

Images