Finger fixing and training control method, device and system of hand training instrument

Abstract

The invention relates to the technical field of medical instruments, and discloses a finger fixing and training control method, device and system for a hand training instrument.The finger fixing method comprises the steps that finger rings and corresponding accessories are printed in a 3D mode, the finger rings are assembled to the positions, corresponding to all fingers, of the hand training instrument, and the fingers of a used individual are fixed through direct drive of a motor; the correct training movement force line and the accurate joint movement range measurement are ensured; the training control method comprises the steps that training parameters are set according to the limit bending and stretching angle of an individual in combination with actual experience data, the individual is guided to be used for training based on a correct training movement force line within the range limited by the training parameters, and a sensor collects various data in the training process in real time and uploads the data to a cloud end; training effect analysis is carried out through the analysis prediction model, remote monitoring and optimization are achieved, training parameters of the terminal are remotely set and adjusted according to data fed back by the terminal, and the system is of great significance to hand function rehabilitation training and control before and after traumatic malformation operation.

Description

Technical Field

[0001] The present invention relates to the technical field of medical devices, and in particular to a finger fixing and training control method, device and system of a hand training device. Background Art

[0002] The hand has a wide range of activities. Fine or heavy work must be done by hand. It is an indispensable and important part of human life. Since the hand is an important labor organ, the incidence of hand trauma caused by various reasons is high, often leading to partial or even complete loss of hand function, causing great inconvenience in life and work for patients. The incomplete hand also destroys the integrity of the human body, brings great mental trauma to patients, hinders participation in social labor and social activities, and imposes a heavy burden on the family and society. How to reduce the hand dysfunction of patients and restore their ability to live and work as soon as possible is particularly important. With the rapid development of modern plastic surgery, hand surgery and microsurgery technology, a good foundation has been laid for the treatment of various traumatic hand deformities. A large number of clinical practices have shown that the correct surgical plan has laid a good foundation for the restoration of hand function in patients with various hand deformities. Surgery alone cannot solve all hand function problems. Accurate rehabilitation training plans before and after surgery also play a very important role. Only by combining rehabilitation treatment with surgery can patients restore their ideal hand function.

[0003] At present, the research focus of domestic and foreign scholars is mostly on applying hand rehabilitation robots to hand dysfunction caused by neurological diseases such as stroke. They usually have a fixed shape to help patients exercise their hand muscles and nerves. However, there are few reports on the application of hand rehabilitation robots before and after traumatic deformity surgery.

[0004] After traumatic deformity surgery, the hands before surgery have problems such as irregular shape and skin graft contracture. Most of the hand training devices on the market cannot effectively fix the deformed hands. If the hand training devices cannot be effectively fixed, the movement force line is easy to shift during rehabilitation training. In addition, for limb scar contracture, through skin grafting and flap surgery, if the newly transplanted skin is not subjected to appropriate flexion and extension training, the skin will soon contract. If the patient does not rely on hand training devices for self-training, he or she will usually not be able to complete the prescribed amount of training due to pain and various reasons. Due to the large number of patients who need rehabilitation, the plastic surgery rehabilitation room can only conduct a few targeted teaching-assisted trainings. Most patients need to return home to persist in medium- and long-term rehabilitation exercises. However, due to the differences in rehabilitation training levels between regions and the inconvenience of communication and guidance between doctors and patients, many patients have not been able to obtain sufficient and effective rehabilitation training, resulting in the inability to further improve the hand and limb functions, and even partial or complete loss of the hand functions that have been acquired, which is very regrettable.

[0005] Therefore, it is of great significance to develop a set of hand function rehabilitation training and control system suitable for pre- and post-operative trauma deformity surgery. This will play a very important role in the remote rehabilitation training and evaluation of patients before and after surgery and in restoring good hand function. Summary of the invention

[0006] The purpose of the present invention is to solve the shortcomings of the above-mentioned prior art and to provide a finger fixation and training control method, device and system for a hand training device to assist users in hand rehabilitation training before and after traumatic deformity surgery.

[0007] On the one hand, a finger fixing and training control method of a hand training device is provided, wherein: The finger fixing method comprises: Based on the problems of each individual hand before and after the trauma deformity surgery, including irregular shape and skin graft contracture, the hand data of the individual is obtained, and the finger ring and corresponding accessories suitable for the current individual hand are 3D printed. The finger ring is assembled to the position corresponding to each finger of the hand training device, and the fingers of the individual are fixed by motor direct drive to ensure the correct training movement force line and accurate joint range of motion measurement; The training control method comprises: The terminal collects the extreme flexion and extension angles of the hand joints of the individual through a number of sensors, sets training parameters based on the individual's extreme flexion and extension angles combined with actual experience data, guides the individual to train based on the correct training motion force line within the range of training parameter restrictions, and the sensors collect various data in real time during the training process; The terminal uploads the collected training data to the cloud, analyzes the training effect through analysis and prediction models, realizes remote monitoring and optimization, and remotely sets and adjusts the training parameters of the terminal based on the data fed back by the terminal.

[0008] Furthermore, the collecting of the individual's hand joint limit flexion and extension angles by using a plurality of sensors further includes: A strain gauge sensor and a photoelectric encoder are arranged on the actuator shaft of the hand training device, and the limit flexion and extension angles are accurately collected by the strain gauge sensor and the photoelectric encoder, wherein: The tension during the maximum extension movement and the pressure during the maximum flexion movement of the individual hand are calculated by measuring the resistance change of the strain gauge sensor, and the extreme flexion and extension angle of the joint is calculated by combining the pressure and the tension; The photoelectric encoder and the actuator shaft of the hand training device are coaxially connected. The photoelectric encoder includes several angle intervals, each interval corresponds to a coding generation comparison table, and the rotation angle of the actuator shaft of the hand training device after the extreme flexion and extension movement is determined by searching the pre-stored comparison table, and the extreme flexion and extension angle of the hand joint is converted.

[0009] Furthermore, the training includes a passive mode, an assist mode and an autonomous mode, specifically including: The passive mode is applicable to the case where the movement of the individual's hand is completely restricted, and the training device controls the individual's hand to flex and extend according to the set training parameters; The power-assist mode is suitable for situations where the movement of the individual's hand is partially restricted, and the training device applies power when the individual is flexing and extending so that the training parameter requirements are met; The autonomous mode is suitable for a situation where the individual's hand movements are completely unrestricted, and the individual controls the training device to perform training according to the set training parameters.

[0010] Furthermore, the various data collected by the sensor in real time during the training process further include: The hand joint activity is collected by the photoelectric encoder, and the hand function level is evaluated by comparing the real-time joint activity with the normal joint activity; The pressure sensor is used to collect the hand force conditions of the individual during training, and the vital sign sensor is used to collect the heart rate and body temperature of the individual. The physical condition of the individual during training is monitored based on the hand force conditions, heart rate and body temperature.

[0011] Furthermore, the training effect analysis by analyzing the prediction model further includes: Preprocess the data received from the cloud server, remove noise and outliers through data cleaning, and then extract features related to the training effect from the cleaned data, including calculating the average pressure, pressure peak and pressure change rate of the fingers under different movements from the pressure data, and extracting the amplitude characteristics of the finger movement and the range of flexion and extension angles from the joint range of motion data; Building an analysis and prediction model based on a machine learning algorithm, and inputting the preprocessed data into the analysis and prediction model for training; The newly fed back sample data is input into the analysis and prediction model for analysis and prediction, so as to determine the individual's rehabilitation status and hand function level and predict future rehabilitation trends.

[0012] Preferably, the machine learning algorithm uses a linear regression model to construct an analysis and prediction model, and the model expression is as follows: in, is the regression coefficient of the model, For the extracted features, the least squares method is used to solve the sum of square errors between the predicted value and the true value, which is shown as follows: Where n is the number of samples in the training data. The training effect value of the i-th sample predicted by the model; The preprocessed data is input into the analysis and prediction model for training, and the performance of the model is evaluated by cross-validation method, including calculating the mean square error and determination coefficient. The prediction value sequence is set as , the true value sequence is , then the calculation formula of mean square error is as follows: The formula for calculating the coefficient of determination is as follows: in, is the mean of the true values.

[0013] Preferably, the remote monitoring and optimization further comprises: The server generates intuitive charts and reports based on the output results of the analysis and prediction model, and displays them through the monitoring platform; An early warning mechanism is set up on the monitoring platform. If the risk of sports injury of an individual user exceeds a preset threshold, an early warning message is automatically sent to the monitoring platform for timely intervention and adjustment.

[0014] Furthermore, the remote setting and adjustment of the training parameters of the terminal according to the data fed back by the terminal using an optimization algorithm further includes: The training intensity and frequency are optimized by the gradient descent algorithm, and the iterative update formula is as follows: in, is the terminal training parameter vector, is the learning rate, which controls the step size of each parameter update. is the objective function at the current parameter value The gradient vector at , by continuously iteratively updating the parameters, makes the objective function gradually converge to the minimum value, thus obtaining the optimal training parameter setting; The optimized and adjusted training parameters are sent to the corresponding terminal device through the network communication protocol. After receiving the new parameters, the terminal automatically updates the settings of the training parameters.

[0015] Furthermore, the training also includes constructing a training environment through virtual reality technology: Customize virtual scenes and rehabilitation games according to the specific condition and rehabilitation goals of the individual, set a number of training actions and tasks in combination with reality, and allow the individual to train through an immersive experience; The sensors capture the movements of the individual, record the movement trajectory, speed, strength and other parameters, compare the collected data with the preset standards, and provide real-time feedback and sharing to the cloud server; Set up a reward mechanism to give corresponding rewards when individuals complete preset training tasks or reach preset exercise indicators, thereby motivating individuals to train.

[0016] On the other hand, a finger fixing and training control device for a hand training device is provided, including a terminal hand training device, a cloud server, a monitoring platform and a virtual reality device: The terminal hand training device includes a wearable manipulator, a drive motor, several sensors, a 3D printed ring, ring accessories, an adjustment component and a microcontroller unit MCU, wherein: The wearable manipulator is electrically connected to the drive motor, and the drive motor provides power for the terminal hand training device; The 3D printed finger ring and the finger ring accessory are assembled on the finger part of the wearable manipulator, and the finger is effectively fixed by the driving motor to achieve the guidance of the correct direction of the motion force line. According to the situation of each finger, each of the 3D printed finger rings is fixed at an adapted position by the finger ring accessory; The adjustment assembly includes an adjustment track, a locking screw and a limiting rod, the limiting rod includes a plurality of holes, the locking screw is assembled in the hole position according to the length requirement of the finger part, and the locking screw is fixed in the position of the adjustment track; The sensors are installed at the joints of the wearable manipulator, and the sensors include strain gauge sensors, photoelectric encoders, pressure sensors and vital sign sensors. The photoelectric encoders and strain gauge sensors are installed on the actuator shaft of the wearable manipulator, and the photoelectric encoders are coaxially connected to the actuator shaft. The terminal hand training device has the micro control unit MCU built in, and the micro control unit MCU is connected to each of the sensors to obtain data collected by the sensors, and transmit the data to the cloud server through a network communication protocol; The monitoring platform is connected to the cloud server and transmits data via a network communication protocol. The monitoring platform is used for data display and early warning alarms. The virtual reality device is connected to the terminal hand training device to provide a virtual training environment construction function for the terminal hand training device.

[0017] In addition, a finger fixing and training control system of a hand training device is provided, including a fixing module and a training control module: The fixing module is used to obtain the hand data of the user, 3D print the finger ring and corresponding accessories, assemble the finger ring to the position corresponding to each finger of the hand training device, and fix the finger of the user through the motor direct drive to ensure the correct training movement force line and accurate joint activity measurement; The training control module includes a collection setting unit and a remote control unit, wherein: The acquisition setting unit is used to collect the extreme flexion and extension angles of the hand joints of the user through a plurality of sensors at the terminal, set the training parameters according to the extreme flexion and extension angles of the user in combination with actual experience data, guide the user to perform training based on the correct training motion force line within the range limited by the training parameters, and the sensors collect various data in real time during the training process; The remote control unit is used for the terminal to upload the collected training data to the cloud, analyze the training effect through the analysis prediction model, realize remote monitoring and optimization, and remotely set and adjust the training parameters of the terminal according to the data fed back by the terminal.

[0018] Compared with the prior art, the present invention has the following beneficial effects: The present invention uses 3D printed finger rings and corresponding accessories to assemble the finger rings to the positions corresponding to each finger of the hand training device, and uses motor direct drive to fix and use individual fingers to ensure the correct training motion force line and accurate joint activity measurement; The present invention collects training data and individual body data by setting various sensors to set training parameters and feedback training results, which is convenient for doctors to control and monitor; The present invention uploads the evaluation data to the cloud and analyzes the training effect through the analysis prediction model to achieve remote monitoring and optimization; The present invention constructs a training environment through virtual reality technology, customizes virtual scenes and rehabilitation games according to the specific condition and rehabilitation goals of the individual, sets a number of training actions and tasks in combination with reality, and allows the individual to be trained through an immersive experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 A flow chart of a training control method of a hand training device according to the present invention; Figure 2 This is a schematic diagram of the structure of a wearable manipulator finger part equipped with a 3D printed ring according to the present invention; Figure 3 This is a schematic diagram of a length adjustment structure of a wearable manipulator finger portion of the present invention; Figure 4 This is a structural block diagram of a finger fixation and training control system module of a hand training device of the present invention.

[0020] Reference numerals 1: Wearable robotic finger part; 2: 3D printed ring; 3: Ring accessories; 31: spiral adjustment rod; 32: finger ring fixing part; 4: Adjust the track; 5: Positioning screw; 6: Limit rod. DETAILED DESCRIPTION

[0021] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0022] Those skilled in the art will appreciate that, unless otherwise stated, the singular forms "a", "an", "said" and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0023] The specific implementation of the present invention is described below with reference to the accompanying drawings and embodiments.

[0024] Example 1 This embodiment provides a finger fixing and training control method for a hand training device, and the technical solution includes a finger fixing method and a training control method, wherein: The finger fixing method comprises: Based on the problems of each individual hand before and after the trauma deformity surgery, including irregular shape and skin graft contracture, the hand data of the individual is obtained, and the finger ring and corresponding accessories suitable for the current individual hand are 3D printed. The finger ring is assembled to the position corresponding to each finger of the hand training device, and the fingers of the individual are fixed by motor direct drive to ensure the correct training movement force line and accurate joint range of motion measurement; The training control method is as follows Figure 1 As shown, including: The terminal collects the extreme flexion and extension angles of the hand joints of the individual through a number of sensors, sets training parameters based on the individual's extreme flexion and extension angles combined with actual experience data, guides the individual to train based on the correct training motion force line within the range of training parameter restrictions, and the sensors collect various data in real time during the training process; The terminal uploads the collected training data to the cloud, analyzes the training effect through analysis and prediction models, realizes remote monitoring and optimization, and remotely sets and adjusts the training parameters of the terminal based on the data fed back by the terminal.

[0025] Wherein, the collecting of the individual's hand joint limit flexion and extension angles by using a plurality of sensors further includes: A strain gauge sensor and a photoelectric encoder are arranged on the actuator shaft of the hand training device, and the limit flexion and extension angles are accurately collected by the strain gauge sensor and the photoelectric encoder, wherein: The tension during the maximum extension movement and the pressure during the maximum flexion movement of the individual hand are calculated by measuring the resistance change of the strain gauge sensor, and the extreme flexion and extension angle of the joint is calculated by combining the pressure and the tension; The photoelectric encoder and the actuator shaft of the hand training device are coaxially connected. The photoelectric encoder includes several angle intervals, each interval corresponds to a coding generation comparison table, and the rotation angle of the actuator shaft of the hand training device after the extreme flexion and extension movement is determined by searching the pre-stored comparison table, and the extreme flexion and extension angle of the hand joint is converted.

[0026] In this embodiment, the training includes a passive mode, an assist mode and an autonomous mode, specifically including: The passive mode is applicable to the case where the movement of the individual's hand is completely restricted, and the training device controls the individual's hand to flex and extend according to the set training parameters; The power-assist mode is suitable for situations where the movement of the individual's hand is partially restricted, and the training device applies power when the individual is flexing and extending so that the training parameter requirements are met; The autonomous mode is suitable for a situation where the individual's hand movements are completely unrestricted, and the individual controls the training device to perform training according to the set training parameters.

[0027] Furthermore, the sensor collects various data in real time during the training process, including: The hand joint activity is collected by the photoelectric encoder, and the hand function level is evaluated by comparing the real-time joint activity with the normal joint activity; The pressure sensor is used to collect the hand force conditions of the individual during training, and the vital sign sensor is used to collect the heart rate and body temperature of the individual. The physical condition of the individual during training is monitored based on the hand force conditions, heart rate and body temperature.

[0028] Among them, existing vital signs sensors include body temperature sensors, respiratory sensors, blood pressure sensors, etc. Equipment that monitors basic life characteristics data of the human body plays a vital role in many fields such as medical care and health monitoring.

[0029] In this embodiment, the process of collecting the range of motion of the joints by the photoelectric encoder is the same as the method used in collecting the extreme flexion and extension angles of the hand joints of an individual as described above.

[0030] Specifically, the collection of the joint range of motion includes: When the joint rotates, the angle change is obtained through the angle sensor, and the joint range of motion is calculated according to the flexion and extension limitation of the user: If the joint can move in a single direction and is limited, the range of motion of the joint is calculated based on the angle of movement in the normal direction and the angle of movement in the limited direction after being limited; If the joint is movable in both directions and is limited, the range of motion of the joint is calculated based on the angles of the restricted two-way motion.

[0031] If a joint moves in only one direction (such as pure flexion-extension) and has no restriction of motion, the range of motion (ROM) is calculated as: ROM = flexion angle - extension angle.

[0032] When a joint's movement in one direction is restricted: If the extension is limited, ROM = normal flexion angle - extension angle after limitation; If flexion is limited, ROM = limited flexion angle - normal extension angle.

[0033] When a joint is restricted in both flexion and extension: ROM = restricted flexion angle - restricted extension angle.

[0034] Furthermore, the training effect analysis by analyzing the prediction model includes: Preprocess the data received from the cloud server, remove noise and outliers through data cleaning, and then extract features related to the training effect from the cleaned data, including calculating the average pressure, pressure peak and pressure change rate of the fingers under different movements from the pressure data, and extracting the amplitude characteristics of the finger movement and the range of flexion and extension angles from the joint range of motion data; Building an analysis and prediction model based on a machine learning algorithm, and inputting the preprocessed data into the analysis and prediction model for training; The newly fed back sample data is input into the analysis and prediction model for analysis and prediction, so as to determine the individual's rehabilitation status and hand function level and predict future rehabilitation trends.

[0035] Specifically, in this embodiment, noise and outliers are removed by data cleaning, and a moving average filtering algorithm is used. Assume that the original data sequence is , the sequence after moving average filtering is , and its calculation formula is: Where m is the size of the moving average window. This algorithm smoothes the data by calculating the average of adjacent data to reduce the impact of noise.

[0036] For outlier detection, we use a statistical-based approach to calculate the mean of the data and standard deviation , will exceed The data in the range are regarded as outliers and corrected or deleted, where k is a threshold set based on experience and is 3 by default.

[0037] In this embodiment, the machine learning algorithm uses a linear regression model to construct an analysis and prediction model, and the model expression is as follows: in, is the regression coefficient of the model, For the extracted features, the least squares method is used to solve the sum of square errors between the predicted value and the true value, which is shown as follows: Where n is the number of samples in the training data. The training effect value of the i-th sample predicted by the model; The preprocessed data is input into the analysis and prediction model for training, and the performance of the model is evaluated by cross-validation method, including calculating the mean square error and determination coefficient. The prediction value sequence is set as , the true value sequence is , then the calculation formula of mean square error is as follows: The formula for calculating the coefficient of determination is as follows: in, is the mean of the true values.

[0038] At the same time, the neural network model can also be used to comprehensively analyze the patient's movement posture data and movement force data to predict the patient's rehabilitation progress and possible sports injury risks. The model is trained based on a large amount of historical rehabilitation training data to learn the characteristic patterns of different rehabilitation training stages and the relationship between them and rehabilitation effects.

[0039] In this embodiment, the results of rehabilitation training include the percentage of rehabilitation progress, the accuracy and standardization scores of rehabilitation movements, and predictions of future rehabilitation trends (such as the estimated time to reach the next rehabilitation stage, training focuses that may need to be adjusted, etc.).

[0040] In addition, the remote monitoring and optimization implementation includes: The server generates intuitive charts and reports based on the output results of the analysis and prediction model, and displays them through the monitoring platform; An early warning mechanism is set up on the monitoring platform. If the risk of sports injury of an individual user exceeds a preset threshold, an early warning message is automatically sent to the monitoring platform for timely intervention and adjustment.

[0041] Specifically, in this embodiment, the intuitive charts and reports include line graphs showing changes in rehabilitation progress over time and bar graphs comparing the completion quality of different training movements.

[0042] Based on the cloud analysis results and remote monitoring conditions, rehabilitation therapists or doctors can remotely set and adjust the training parameters of the terminal device directly on the monitoring platform.

[0043] Furthermore, the remote setting and adjustment of the training parameters of the terminal according to the data fed back by the terminal using an optimization algorithm further includes: The training intensity and frequency are optimized by the gradient descent algorithm, and the iterative update formula is as follows: in, is the terminal training parameter vector, is the learning rate, which controls the step size of each parameter update. is the objective function at the current parameter value The gradient vector at , by continuously iteratively updating the parameters, makes the objective function gradually converge to the minimum value, thus obtaining the optimal training parameter setting; The optimized and adjusted training parameters are sent to the corresponding terminal device through the network communication protocol. After receiving the new parameters, the terminal automatically updates the settings of the training parameters.

[0044] In this embodiment, if the analysis and prediction model determines that the user individual is not sufficiently trained in strength in a certain training action, the resistance parameter and the number of training times are increased remotely; If the individual's joint mobility recovery effect is excellent, the difficulty level of the training movements should be appropriately increased, including increasing the flexion and extension angle range and flexion and extension speed.

[0045] Through such a complete cloud-based remote monitoring and optimization system, the efficiency and quality of sports rehabilitation training can be greatly improved, making it easier for patients to carry out effective rehabilitation training at home or in other non-medical institutions. It also makes it easier for rehabilitation professionals to track and provide personalized guidance to patients throughout the process.

[0046] Furthermore, the training also includes constructing a training environment through virtual reality technology: Customize virtual scenes and rehabilitation games according to the specific condition and rehabilitation goals of the individual, set a number of training actions and tasks in combination with reality, and allow the individual to train through an immersive experience; The sensors capture the movements of the individual, record the movement trajectory, speed, strength and other parameters, compare the collected data with the preset standards, and provide real-time feedback and sharing to the cloud server; Set up a reward mechanism to give corresponding rewards when individuals complete preset training tasks or reach preset exercise indicators, thereby motivating individuals to train.

[0047] Specifically, VR technology can be used to create a variety of realistic sports scenes. It can simulate the actions of picking up and placing items in a virtual kitchen. Patients seem to be in a real kitchen environment, and according to the system prompts, they can grab different kitchen utensils, such as picking up a frying pan and putting it on the stove. Such tasks can effectively exercise the strength, coordination and joint mobility of the upper limbs.

[0048] In this embodiment, the feedback can be visual, such as using different colored lines in the VR screen to show the difference between the patient's actual movement trajectory and the ideal movement trajectory; it can also be auditory, when the patient's movements are not standard, voice prompts are used to correct the movements, such as prompting "Your arm is not raised enough, please raise it higher." This real-time feedback allows patients to immediately understand their training status, helps to quickly adjust movements, and improve the effect of rehabilitation training.

[0049] The reward mechanism set up in this implementation is that after the patient completes the required number of hand flexion and extension training within a week, he can unlock a new game scene, such as exploring mysterious interstellar ruins, thereby encouraging the patient to continue rehabilitation training.

[0050] In addition, multi-person virtual rehabilitation scenes can be created to allow patients and therapists or patients to conduct interactive rehabilitation training. For example, in a virtual hand rehabilitation training scene, the therapist can lead the patient to complete the movements together, and the patient can better imitate the therapist's standard movements, which also increases the fun of rehabilitation training.

[0051] In addition, this embodiment provides a finger fixing and training control device for a hand training device, which is used to implement the above method, including a terminal hand training device, a cloud server, a monitoring platform and a virtual reality device: The terminal hand training device includes a wearable manipulator, a drive motor, several sensors, a 3D printed finger ring 2, a finger ring accessory 3, an adjustment component and a microcontroller unit MCU, wherein: The wearable manipulator is electrically connected to the drive motor, and the drive motor provides power for the terminal hand training device; like Figure 2 As shown, the 3D printed finger ring 2 and the finger ring accessory 3 are assembled on the wearable manipulator finger part 1, and the finger is effectively fixed by the driving motor to achieve the guidance of the correct direction of the motion force line. According to the situation of each finger, each of the 3D printed finger rings 2 is fixed at an adapted position by the finger ring accessory 3. Specifically, in this embodiment, the finger ring accessory 3 includes a spiral adjustment rod 31 and a finger ring fixing component 32, and the spiral adjustment rod 31 includes a threaded rod. The upper part of the 3D printed finger ring 2 is fixed to the spiral adjustment rod 31 by the finger ring fixing component 32, wherein the fixing position on the spiral adjustment rod 31 is flexibly selected according to the situation of each finger of the patient; Secondly, if Figure 3 As shown, the adjustment assembly includes an adjustment rail 4, a locking screw 5 and a limiting rod 6. The limiting rod 6 includes a plurality of holes. The locking screw 5 is assembled in the hole position according to the length requirement of the finger part, and the locking screw 5 is fixed in the position of the adjustment rail 4 by the locking screw 5; The sensors are installed at the joints of the wearable manipulator, and the sensors include strain gauge sensors, photoelectric encoders, pressure sensors and vital sign sensors. The photoelectric encoders and strain gauge sensors are installed on the actuator shaft of the wearable manipulator, and the photoelectric encoders are coaxially connected to the actuator shaft. The terminal hand training device has the micro control unit MCU built in, and the micro control unit MCU is connected to each of the sensors to obtain data collected by the sensors, and transmit the data to the cloud server through a network communication protocol; The monitoring platform is connected to the cloud server and transmits data via a network communication protocol. The monitoring platform is used for data display and early warning alarms. The virtual reality device is connected to the terminal hand training device to provide a virtual training environment construction function for the terminal hand training device.

[0052] In addition, this embodiment also provides a finger fixing and training control system for a hand training device, such as Figure 4 As shown, it includes a fixed module and a training control module: The fixing module is used to obtain the hand data of the user, 3D print the finger ring and corresponding accessories, assemble the finger ring to the position corresponding to each finger of the hand training device, and fix the finger of the user through the motor direct drive to ensure the correct training movement force line and accurate joint activity measurement; The training control module includes a collection setting unit and a remote control unit, wherein: The acquisition setting unit is used to collect the extreme flexion and extension angles of the hand joints of the user through a plurality of sensors at the terminal, set the training parameters according to the extreme flexion and extension angles of the user in combination with actual experience data, guide the user to perform training based on the correct training motion force line within the range limited by the training parameters, and the sensors collect various data in real time during the training process; The remote control unit is used for the terminal to upload the collected training data to the cloud, analyze the training effect through the analysis prediction model, realize remote monitoring and optimization, and remotely set and adjust the training parameters of the terminal according to the data fed back by the terminal.

[0053] in, Figure 3 The functional implementation of each module and unit in the finger fixation and training control system of the hand training device corresponds to the steps in the above-mentioned finger fixation and training control method embodiment of the hand training device, and its functions and implementation processes will not be repeated here one by one.

[0054] Finally, it should be noted that the above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

[0055] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A finger fixing and training control method for a hand training device, characterized in that: in, The finger fixing method comprises: Based on the problems of each individual hand before and after the trauma deformity surgery, including irregular shape and skin graft contracture, the hand data of the individual is obtained, and the finger ring and corresponding accessories suitable for the current individual hand are 3D printed. The finger ring is assembled to the position corresponding to each finger of the hand training device, and the fingers of the individual are fixed by motor direct drive to ensure the correct training movement force line and accurate joint range of motion measurement; The training control method comprises: The terminal collects the extreme flexion and extension angles of the hand joints of the individual through a number of sensors, sets training parameters based on the individual's extreme flexion and extension angles combined with actual experience data, guides the individual to train based on the correct training motion force line within the range of training parameter restrictions, and the sensors collect various data in real time during the training process; The terminal uploads the collected training data to the cloud, analyzes the training effect through analysis and prediction models, realizes remote monitoring and optimization, and remotely sets and adjusts the training parameters of the terminal based on the data fed back by the terminal.

2. The finger fixing and training control method of the hand training device according to claim 1, characterized in that: The collecting of the individual's hand joint limit flexion and extension angles by using a plurality of sensors further includes: A strain gauge sensor and a photoelectric encoder are arranged on the actuator shaft of the hand training device, and the limit flexion and extension angles are accurately collected by the strain gauge sensor and the photoelectric encoder, wherein: The tension during the maximum extension movement and the pressure during the maximum flexion movement of the individual hand are calculated by measuring the resistance change of the strain gauge sensor, and the extreme flexion and extension angle of the joint is calculated by combining the pressure and the tension; The photoelectric encoder and the actuator shaft of the hand training device are coaxially connected. The photoelectric encoder includes several angle intervals, each interval corresponds to a coding generation comparison table, and the rotation angle of the actuator shaft of the hand training device after the extreme flexion and extension movement is determined by searching the pre-stored comparison table, and the extreme flexion and extension angle of the hand joint is converted.

3. The finger fixing and training control method of the hand training device according to claim 1, characterized in that: The training includes passive mode, assisted mode and autonomous mode, specifically including: The passive mode is applicable to the case where the movement of the individual's hand is completely restricted, and the training device controls the individual's hand to flex and extend according to the set training parameters; The power-assist mode is suitable for situations where the movement of the individual's hand is partially restricted, and the training device applies power when the individual is flexing and extending so that the training parameter requirements are met; The autonomous mode is suitable for a situation where the individual's hand movements are completely unrestricted, and the individual controls the training device to perform training according to the set training parameters.

4. The finger fixing and training control method of the hand training device according to claim 2, characterized in that: The various data collected by the sensor in real time during the training process further include: The hand joint activity is collected by the photoelectric encoder, and the hand function level is evaluated by comparing the real-time joint activity with the normal joint activity; The pressure sensor is used to collect the hand force conditions of the individual during training, and the vital sign sensor is used to collect the heart rate and body temperature of the individual. The physical condition of the individual during training is monitored based on the hand force conditions, heart rate and body temperature.

5. The finger fixing and training control method of the hand training device according to claim 4, characterized in that: The training effect analysis through the analysis prediction model further includes: Preprocess the data received from the cloud server, remove noise and outliers through data cleaning, and then extract features related to the training effect from the cleaned data, including calculating the average pressure, pressure peak and pressure change rate of the fingers under different movements from the pressure data, and extracting the amplitude characteristics of the finger movement and the range of flexion and extension angles from the joint range of motion data; Building an analysis and prediction model based on a machine learning algorithm, and inputting the preprocessed data into the analysis and prediction model for training; The newly fed back sample data is input into the analysis and prediction model for analysis and prediction, so as to determine the individual's rehabilitation status and hand function level and predict future rehabilitation trends.

6. The finger fixing and training control method of the hand training device according to claim 5, characterized in that: The machine learning algorithm uses a linear regression model to build an analysis and prediction model. The model expression is as follows: in, is the regression coefficient of the model, For the extracted features, the least squares method is used to solve the sum of square errors between the predicted value and the true value to be minimized, as shown below: Where n is the number of samples in the training data. The training effect value of the i-th sample predicted by the model; The preprocessed data is input into the analysis and prediction model for training, and the performance of the model is evaluated by cross-validation method, including calculating the mean square error and determination coefficient. The prediction value sequence is set as , the true value sequence is , then the calculation formula of mean square error is as follows: The formula for calculating the coefficient of determination is as follows: in, is the mean of the true values.

7. The finger fixing and training control method of the hand training device according to claim 6, characterized in that: The remote monitoring and optimization further comprises: The server generates intuitive charts and reports based on the output results of the analysis and prediction model, and displays them through the monitoring platform; An early warning mechanism is set up on the monitoring platform. If the risk of sports injury of an individual user exceeds a preset threshold, an early warning message is automatically sent to the monitoring platform for timely intervention and adjustment.

8. The finger fixing and training control method of the hand training device according to claim 1, characterized in that: The method of remotely setting and adjusting the training parameters of the terminal according to the data fed back by the terminal using an optimization algorithm further includes: The training intensity and frequency are optimized by the gradient descent algorithm, and the iterative update formula is as follows: in, is the terminal training parameter vector, is the learning rate, which controls the step size of each parameter update. is the objective function at the current parameter value The gradient vector at , by continuously iteratively updating the parameters, makes the objective function gradually converge to the minimum value, thus obtaining the optimal training parameter setting; The optimized and adjusted training parameters are sent to the corresponding terminal device through the network communication protocol. After receiving the new parameters, the terminal automatically updates the settings of the training parameters.

9. The finger fixing and training control method of the hand training device according to claim 1, characterized in that: The training also includes building a training environment through virtual reality technology: Customize virtual scenes and rehabilitation games according to the specific condition and rehabilitation goals of the individual, set a number of training actions and tasks in combination with reality, and allow the individual to train through an immersive experience; The sensors capture the movements of the individual, record the movement trajectory, speed, strength and other parameters, compare the collected data with the preset standards, and provide real-time feedback and sharing to the cloud server; Set up a reward mechanism to give corresponding rewards when individuals complete preset training tasks or reach preset exercise indicators, thereby motivating individuals to train.

10. A finger fixing and training control device of a hand training apparatus for implementing the method described in any one of claims 1 to 9, characterized in that: Including terminal hand training equipment, cloud server, monitoring platform and virtual reality equipment: The terminal hand training device includes a wearable manipulator, a drive motor, several sensors, a 3D printed ring, ring accessories, an adjustment component and a microcontroller unit MCU, wherein: The wearable manipulator is electrically connected to the drive motor, and the drive motor provides power for the terminal hand training device; The 3D printed finger ring and the finger ring accessory are assembled on the finger part of the wearable manipulator, and the finger is effectively fixed by the driving motor to achieve the guidance of the correct direction of the motion force line. According to the situation of each finger, each of the 3D printed finger rings is fixed at an adapted position by the finger ring accessory; The adjustment assembly includes an adjustment track, a locking screw and a limiting rod, the limiting rod includes a plurality of holes, the locking screw is assembled in the hole position according to the length requirement of the finger part, and the locking screw is fixed in the position of the adjustment track; The sensors are installed at the joints of the wearable manipulator, and the sensors include strain gauge sensors, photoelectric encoders, pressure sensors and vital sign sensors. The photoelectric encoders and strain gauge sensors are installed on the actuator shaft of the wearable manipulator, and the photoelectric encoders are coaxially connected to the actuator shaft. The terminal hand training device has the micro control unit MCU built in, and the micro control unit MCU is connected to each of the sensors to obtain data collected by the sensors, and transmit the data to the cloud server through a network communication protocol; The monitoring platform is connected to the cloud server and transmits data via a network communication protocol. The monitoring platform is used for data display and early warning alarms. The virtual reality device is connected to the terminal hand training device to provide a virtual training environment construction function for the terminal hand training device.

11. A finger fixing and training control system for a hand training device, characterized in that: Including fixed module and training control module: The fixing module is used to obtain the hand data of the user, 3D print the finger ring and corresponding accessories, assemble the finger ring to the position corresponding to each finger of the hand training device, and fix the finger of the user through the motor direct drive to ensure the correct training movement force line and accurate joint activity measurement; The training control module includes a collection setting unit and a remote control unit, wherein: The acquisition setting unit is used to collect the extreme flexion and extension angles of the hand joints of the user through a plurality of sensors at the terminal, set the training parameters according to the extreme flexion and extension angles of the user in combination with actual experience data, guide the user to perform training based on the correct training motion force line within the range limited by the training parameters, and the sensors collect various data in real time during the training process; The remote control unit is used for the terminal to upload the collected training data to the cloud, analyze the training effect through the analysis prediction model, realize remote monitoring and optimization, and remotely set and adjust the training parameters of the terminal according to the data fed back by the terminal.

Images