Multi-mode intelligent interactive life function auxiliary system for patient with hemiplegic upper limbs

Abstract

The invention provides a multi-mode intelligent interactive life function auxiliary system for patients with upper limb hemiplegia, which relates to the field of life function auxiliary systems for patients with upper limb hemiplegia, and is characterized by comprising a mechanical body structure, a driving and sensing system, a touch and sensing system and control and training software, the mechanical body structure comprises modular exoskeleton structures covering shoulder joints, elbow joints, forearms / wrist joints and finger joints, the driving and sensing system comprises a lightweight motor-cable transmission mechanism, torque output and compliance are achieved by combining a speed reducer and an elastic element, and torque sensors and position encoders are integrated on all the joints. The system has the advantages that a patient can obtain tactile perception in the rehabilitation process through a driving and sensing system, the motion participation sense is enhanced, the system is further integrated with a rehabilitation evaluation module, hand flexibility, response speed and recovery speed progress are quantitatively scored based on a machine learning algorithm, and a training suggestion report is output for a doctor.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of life function auxiliary system for upper limb hemiplegic patients, and particularly relates to a multi-modal intelligent interactive life function auxiliary system for upper limb hemiplegic patients. BACKGROUND

[0002] With the aggravation of population aging and the rising incidence of stroke, the number of patients with upper limb motor dysfunction continues to grow. As an important auxiliary treatment method, rehabilitation robots have gradually attracted attention. Wearable upper limb rehabilitation mechanical arms, as a leading technology for assisting hemiplegic patients in functional recovery, have received widespread attention worldwide. As early as the 1960s, exoskeleton-type assistive devices began to emerge. After several technical iterations, in recent years, with the progress of materials, sensing and control technologies, upper limb exoskeletons have entered a new stage of development. Various rehabilitation robots for upper limb paralysis have been introduced internationally.

[0003] At present, the existing upper limb rehabilitation robots have many problems. In terms of lightweight design, the traditional rigid link structure leads to a mechanical arm weight of more than 3 kg. The carbon fiber support rod lacks bionic curved surface design. The threaded adjustment takes too long. The joint rotation center deviation causes a sharp increase in shear stress. In terms of drive system, the rigid drive joint stiffness is too high. The force control algorithm response is lagging. There is a lack of elastic energy storage mechanism. In terms of intention recognition, the single signal source recognition accuracy is insufficient. The traditional threshold judgment cannot adapt to time-varying characteristics. There is a lack of personalized adaptive model. In terms of wearing safety, the binding type fixing takes a long time to put on and take off. There is a lack of multi-level safety protection mechanism. The poor material breathability leads to abnormal skin temperature, which causes many inconveniences in use. SUMMARY

[0004] The purpose of the present application is to provide a multi-modal intelligent interactive life function auxiliary system for upper limb hemiplegic patients, which solves the problems in the prior art

[0005] In order to achieve the above-mentioned purpose of the application, the technical scheme adopted by the present application is:

[0006] The application discloses a multi-modal intelligent interactive life function auxiliary system for upper limb hemiplegic patients, which is characterized by comprising a mechanical body structure, a driving and sensing system, a tactile and sensing system and a control and training software, wherein the mechanical body structure comprises a modular exoskeleton structure covering shoulder joints, elbow joints, forearm / wrist joints and finger joints; the driving and sensing system comprises a lightweight motor-cable transmission mechanism, a reducer and an elastic element, torque sensors and position encoders integrated in each joint, a bioelectric signal acquisition module arranged at key positions of the upper limbs of the patients for acquiring electromyographic signals (EMG) or electroencephalographic signals (EEG), and a control and training software containing a motion control algorithm, a biofeedback intention recognition control module, a safety control strategy and upper computer software.

[0007] As an improvement, the mechanical body structure adopts a modular design and has six active degrees of freedom, wherein the shoulder joint is configured with three degrees of freedom, the elbow joint is configured with one degree of freedom, the forearm rotation is configured with one degree of freedom, and the hand is configured with one extension and contraction degree of freedom; the structure comprises telescopic connecting rods and slide rail mechanisms to adapt to patients with different body types, and quick fixing devices are arranged to ensure wearing safety; the three degrees of freedom of the shoulder joint of the mechanical body structure comprise a flexion and extension degree of freedom, an abduction / adduction degree of freedom and a pronation / supination degree of freedom.

[0008] As an improvement, the tactile and sensing system integrates multi-point pressure sensors and micro vibration feedback devices in the hand and fingertip areas, a rehabilitation evaluation module quantifies hand flexibility, reaction speed and recovery progress based on a machine learning algorithm, dynamically adjusts training parameters and generates an individualized training path.

[0009] As an improvement, the control and training software supports passive training, power assistance control and resistance training modes, an intention recognition control module triggered by weak electromyographic / encephalographic signals, a safety control strategy torque limit, an emergency stop mechanism and an abnormal unloading protection, and upper computer software for providing a graphical interface, a training task library (daily activity simulation, game-based training), supporting data record analysis and trend evaluation.

[0010] As an improvement, the bioelectric signal acquisition module of the driving and sensing system adopts a multi-channel EMG sensor array, is arranged at the triceps brachii and the forearm radial wrist flexor muscle area, and suppresses power frequency interference through a wavelet transform algorithm.

[0011] As an improvement, the rehabilitation evaluation module of the tactile and sensing system contains a finger flexibility scoring unit that calculates the fingertip contact area and pressure distribution based on pressure sensor data, a motion trajectory analysis unit that can evaluate joint range of motion in combination with encoder data, and a dynamic parameter adjustment unit that can adaptively adjust the vibration feedback intensity and resistance level according to the scoring results.

[0012] As an improvement, the safety control strategy of the control and training software uses a fuzzy PID algorithm to adjust the output torque in real time, and triggers the robot arm to quickly reset to a zero-gravity state when the joint torque exceeds the preset threshold.

[0013] As an improvement, the task library of the upper computer software contains a virtual reality (VR) hand grip simulation scene, a fine operation game developed based on Unity3D, and a training data visualization panel that can display the electromyographic signal intensity, joint movement angle, and force feedback curve in real time.

[0014] The beneficial effects of the present application are:

[0015] 1、In the present application, the mechanical body structure can be designed for hemiplegic patients, as hemiplegic patients usually have limited hand and upper arm movement function, especially finger function recovery is often more difficult, the device integrates hand exoskeleton and upper arm support structure, which can assist the rehabilitation training of finger fine action and large range of upper arm movement at the same time, thereby emphasizing the recovery of finger function while also considering the improvement of upper arm movement ability, realizing the coordinated rehabilitation treatment of the overall movement function of upper limbs.

[0016] 2、In the present application, the mechanical body structure is designed, and the mechanical arm can adapt to patients of different body shapes, and patients can independently wear and use at home, which will create a new mode of transformation from "hospital equipment" to "portable equipment at home" for domestic upper limb rehabilitation robots, greatly improving the convenience and frequency of rehabilitation training.

[0017] 3、In the present application, through the driving and sensing system, patients can obtain tactile perception during rehabilitation, and the system will also integrate a rehabilitation evaluation module, which can quantitatively score hand flexibility, reaction speed and recovery speed progress based on machine learning algorithm, and output training recommendation report for doctors.

[0018] 4、In the present application, through the design of the control and training software, patients can trigger the mechanical arm to assist movement by trying active movement (weak electromyographic / EEG signal), realize "intentional assisted movement", enhance active participation, and at the same time, design safety control strategies such as torque limitation and emergency stop mechanism to ensure that the device quickly unloads force to protect the patient in abnormal situations. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1A system architecture schematic diagram of a multi-modal intelligent interactive life function auxiliary system for upper limb hemiplegic patients. DETAILED DESCRIPTION

[0020] In order to make the content of the present application easier to be clearly understood, the technical solutions in the embodiments of the present application will be clearly and completely described below in combination with the drawings in the embodiments of the present application. The same parts are denoted by the same reference numerals. It should be noted that the words "front", "back", "left", "right", "up" and "down" used in the following description refer to the directions in the drawings, and the words "inner" and "outer" refer to the directions towards or away from the geometric center of a particular part.

[0021] As shown in the drawings, a multi-modal intelligent interactive life function auxiliary system for upper limb hemiplegic patients comprises a mechanical body structure, a driving and sensing system, a tactile and sensing system, and a control and training software. Figure 1 The mechanical body structure comprises a modular exoskeleton structure covering shoulder joints, elbow joints, forearm / wrist joints and finger joints. The driving and sensing system comprises a lightweight motor-cable transmission mechanism, which combines a reducer and an elastic element to realize output torque and compliance. Each joint is integrated with a torque sensor and a position encoder for real-time monitoring of the force and angular position of the mechanical arm. The patient's upper limb key parts are arranged with bioelectric signal acquisition modules for obtaining electromyographic signals (EMG) or electroencephalographic signals (EEG). Multi-modal sensing information is fused to identify motion intention. The control and training software includes motion control algorithms, biofeedback intention recognition control modules, safety control strategies and upper computer software.

[0022] The mechanical body structure adopts a modular design and has 6 active degrees of freedom. The shoulder joint is configured with 3 degrees of freedom, the elbow joint is configured with 1 degree of freedom, the forearm rotation is configured with 1 degree of freedom, and the hand is configured with 1 extension and contraction degree of freedom. The structure includes telescopic links and slide rail mechanisms to adapt to patients of different body types, and quick fixing devices are provided to ensure wearing safety. The 3 degrees of freedom of the shoulder joint of the mechanical body structure include flexion / extension, abduction / adduction and internal / external rotation degrees of freedom.

[0023] The tactile and sensing system integrates multi-point pressure sensors and micro-vibration feedback devices in the hand and fingertip areas. The rehabilitation evaluation module quantifies hand flexibility, reaction speed and recovery progress based on machine learning algorithms, dynamically adjusts training parameters and generates personalized training paths.

[0024] The motion control algorithm in the control and training software supports passive training, power-assisted control, and resistance training modes. The biofeedback intention recognition control module triggers the robot arm's assisted movements through weak electromyography / electroencephalography signals. The safety control strategy includes torque limitation, emergency stop mechanism, and abnormal force unloading protection. The host computer software is used to provide a graphical interface and a training task library (daily activity simulation, gamification training), and supports data recording analysis and trend evaluation.

[0025] The bioelectric signal acquisition module of the driving and sensing system adopts a multi-channel EMG sensor array, which is deployed in the deltoid muscle of the upper arm and the radial flexor carpi of the forearm, and suppresses power frequency interference through the wavelet transform algorithm.

[0026] The rehabilitation assessment module of the tactile and sensing system includes a finger flexibility scoring unit that calculates the fingertip contact area and pressure distribution based on pressure sensor data; a motion trajectory analysis unit that can evaluate the range of joint motion in combination with encoder data; and a dynamic parameter adjustment unit that can adaptively adjust the vibration feedback intensity and resistance level according to the scoring results.

[0027] The safety control strategy of the control and training software uses a fuzzy PID algorithm to adjust the output torque in real time. When it detects that the joint torque exceeds the preset threshold, it triggers the robotic arm to quickly reset to a zero-gravity state.

[0028] The task library of the host computer software includes virtual reality (VR) hand grasping simulation scenes, fine operation games developed based on Unity3D, and a training data visualization panel that can display the electromyographic signal intensity, joint movement angle and force feedback curve in real time.

[0029] During use, the patient first dons the modular exoskeleton. A quick-fix mechanism adjusts the shoulder, elbow, forearm, and finger structures to suit their body shape. A multi-channel EMG sensor array is then deployed in the deltoid and radial flexor carpi area. After the system is activated, a wavelet transform algorithm is used to suppress power frequency interference and collect EMG signals in real time. The therapist selects a training mode (passive, assisted, or resisted) via host computer software. The patient uses weak EMG signals to trigger robotic arm movements. The system integrates real-time data from torque sensors and position encoders, dynamically adjusting output torque using a fuzzy PID algorithm to ensure movement within safe thresholds. The tactile system's finger pressure sensors monitor grasping movements and calculate a flexibility score based on pressure distribution. The vibration feedback device adaptively adjusts intensity based on the score. The rehabilitation assessment module synchronizes encoder trajectory data to optimize training parameters. The host computer displays EMG intensity, joint angles, and force feedback curves in real time. The patient completes daily activity simulations in a VR grasping scenario or a Unity 3D game. Training data is automatically recorded and analyzed to generate a personalized recovery progress report. Abnormal torque triggers an emergency stop and reset. After training, the system generates multimodal data trend charts to assist in efficacy evaluation.

[0030] The above merely describes preferred embodiments of the present application, and is not used to limit the present application, any modification, equivalent replacement and improvement made within the spirit and principle of the present application should be included in the protection scope of the present application.

Claims

1. A multimodal intelligent interactive life function assistance system for patients with upper limb hemiplegia, characterized by: include: The mechanical body structure, drive and sensing system, tactile and sensing system and control and training software are as follows: the mechanical body structure includes a modular exoskeleton structure covering the shoulder joint, elbow joint, forearm / wrist joint and finger joint; the drive and sensing system includes a lightweight motor-cable transmission mechanism, combined with a reducer and elastic elements to achieve output torque and compliance; each joint is integrated with a torque sensor and a position encoder for real-time monitoring of the force and angular position of the robotic arm; each joint is integrated with a torque sensor and a position encoder for real-time monitoring of the force and angular position of the robotic arm; a bioelectric signal acquisition module is arranged at key parts of the patient's upper limb to obtain electromyographic signals (EMG) or electroencephalogram (EEG), and multimodal sensing information is integrated to identify movement intentions; the control and training software includes a motion control algorithm, a biofeedback intention recognition control module, a safety control strategy and a host computer software.

2. A multimodal intelligent interactive life function assistance system for upper limb hemiplegia patients according to claim 1, characterized in that: The mechanical body structure adopts a modular design and has 6 active degrees of freedom, including 3 degrees of freedom in the shoulder joint, 1 degree of freedom in the elbow joint, 1 degree of freedom in forearm rotation, and 1 degree of freedom in telescopic movement of the hand. The structure includes a telescopic connecting rod and a slide rail mechanism to adapt to patients of different body shapes, and a quick fixing device is provided to ensure safe wearing. The 3 degrees of freedom of the shoulder joint of the mechanical body structure include flexion and extension, abduction / adduction, and internal and external rotation.

3. The multimodal intelligent interactive life function assistance system for upper limb hemiplegia patients according to claim 1 is characterized in that: The tactile and sensing system integrates multi-point pressure sensors and micro-vibration feedback devices in the hand and fingertip areas. The rehabilitation assessment module quantifies hand flexibility, reaction speed and recovery progress based on a machine learning algorithm, dynamically adjusts training parameters and generates a personalized training path.

4. The multimodal intelligent interactive life function assistance system for upper limb hemiplegia patients according to claim 1 is characterized in that: The motion control algorithm in the control and training software supports passive training, power-assisted control, and resistance training modes, a biofeedback intention recognition control module, triggers robotic arm assisted movements through weak electromyographic / electroencephalographic signals, and a safety control strategy with torque limitation, emergency stop mechanism, and abnormal force unloading protection. The host computer software is used to provide a graphical interface and a training task library (daily activity simulation, gamification training), and supports data recording analysis and trend evaluation.

5. The multimodal intelligent interactive life function assistance system for upper limb hemiplegia patients according to claim 1 is characterized in that: The bioelectric signal acquisition module of the driving and sensing system adopts a multi-channel EMG sensor array, which is deployed in the deltoid muscle of the upper arm and the radial flexor carpi of the forearm, and suppresses power frequency interference through a wavelet transform algorithm.

6. The multimodal intelligent interactive life function assistance system for upper limb hemiplegia patients according to claim 1, characterized in that: The rehabilitation assessment module of the tactile and sensing system includes a finger flexibility scoring unit that calculates the fingertip contact area and pressure distribution based on pressure sensor data, a motion trajectory analysis unit that can evaluate the joint motion range in combination with encoder data, and a dynamic parameter adjustment unit that can adaptively adjust the vibration feedback intensity and resistance level according to the scoring results.

7. The multimodal intelligent interactive life function assistance system for upper limb hemiplegia patients according to claim 1 is characterized in that: The safety control strategy of the control and training software adopts a fuzzy PID algorithm to adjust the output torque in real time, and triggers the robotic arm to quickly reset to a zero-gravity state when it is detected that the joint torque exceeds a preset threshold.

8. The multimodal intelligent interactive life function assistance system for upper limb hemiplegia patients according to claim 1 is characterized in that: The task library of the host computer software includes virtual reality (VR) hand grasping simulation scenes, fine operation games developed based on Unity3D, and a training data visualization panel that can display the electromyographic signal intensity, joint movement angle and force feedback curve in real time.

Images