Glove device for promoting hand rehabilitation of stroke patient based on intelligent voice interaction

Abstract

The invention discloses a glove device for promoting hand rehabilitation of a stroke patient based on intelligent voice interaction, and relates to the technical field of combination of a medical rehabilitation technology and an artificial intelligence technology. The glove device comprises a glove body, a control module, a data acquisition module, a pickup module and a voice loudspeaker, the glove body is used for being worn on the hand of the stroke patient and assisting the hand in rehabilitation exercise, and the control module, the data acquisition module, the pickup module and the voice loudspeaker are arranged on the glove body. The control module is used for extracting feature data from hand motion state data received in real time and individual basic information data and / or individual historical exercise data of a stroke patient, and then importing the feature data into a hand rehabilitation evaluation model which completes pre-training based on a deep learning algorithm. A personalized hand rehabilitation exercise scheme of the stroke patient is adjusted according to the hand rehabilitation evaluation result, then a voice signal used for introducing the exercise scheme is transmitted to the voice loudspeaker, and finally the glove body is controlled to execute the exercise scheme according to a voice instruction.

Description

Technical Field

[0001] The present invention belongs to the technical field of the combination of medical rehabilitation technology and artificial intelligence technology, and specifically relates to a glove device that promotes hand rehabilitation of stroke patients based on intelligent voice interaction. Background Art

[0002] Stroke, commonly known as a stroke, is classified into two types: ischemic and hemorrhagic. It is caused by various factors that damage cerebral blood vessels, resulting in focal or global brain tissue damage. Traditional methods of hand neurological and motor rehabilitation for stroke patients often rely on manual manipulation and guidance from physical therapists. This approach is not only inefficient but also fails to meet the growing patient needs. Furthermore, physical therapists have limited time and energy, making it difficult to provide consistent and personalized rehabilitation training for each patient.

[0003] While some existing electronic rehabilitation devices can provide repetitive training movements to a certain extent—for example, the existing patent "CN213588950U - A Simple Pneumatic Hand Rehabilitation Device" utilizes components such as a balloon, airbag gloves, and a trachea tube to easily achieve hand extension—they lack intelligence and interactivity. Specifically, they cannot flexibly adjust based on the patient's real-time feedback and rehabilitation progress, resulting in unsatisfactory rehabilitation results. Furthermore, existing rehabilitation devices rarely address voice interaction, leaving patients passively accepting training during the rehabilitation process, unable to actively participate in and control the training process, and lacking motivation and initiative. For stroke patients, hand rehabilitation requires long-term persistence, and monotonous rehabilitation training can easily lead to boredom and reduced compliance. Furthermore, existing rehabilitation devices also have shortcomings in data collection and analysis, including the inability to accurately record patient training data and the difficulty in objectively evaluating and accurately improving rehabilitation outcomes.

[0004] Therefore, how to provide stroke patients with an effective and innovative new hand function recovery solution in terms of intelligence, interactivity, personalization and data utilization, so as to meet the patients' efficient, convenient and personalized rehabilitation needs, help patients improve their hand motor ability and improve their quality of life, is a topic that technical personnel in this field urgently need to study. Summary of the Invention

[0005] The purpose of the present invention is to provide a glove device that promotes hand rehabilitation of stroke patients based on intelligent voice interaction, so as to solve the problem that the existing electronic assisted rehabilitation solutions for hand nerve rehabilitation and motor rehabilitation of stroke patients lack intelligence and interactivity, making it difficult to meet the patients' efficient, convenient and personalized rehabilitation needs and limiting the rehabilitation training effect.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] The present invention provides a glove device for promoting hand rehabilitation of stroke patients based on intelligent voice interaction, comprising a glove body for wearing on the hand of a stroke patient to assist the hand in rehabilitation exercises, and a control module, a data acquisition module, a sound pickup module, and a voice speaker configured on the glove body;

[0008] The data acquisition module is communicatively connected to the control module, and is used to collect the hand movement state data of the stroke patient in real time, and transmit the hand movement state data to the control module in real time;

[0009] The sound pickup module is communicatively connected to the control module and is used to collect voice signals from the stroke patient in real time and transmit the voice signals to the control module in real time;

[0010] The voice speaker is communicatively connected to the control module and is used to play the voice signal from the control module;

[0011] The control module is used to extract feature data from the hand movement state data received in real time during the completion process and the individual basic information data and / or individual historical exercise data of the stroke patient after completing at least one set of hand rehabilitation exercise movements each time, and then import the feature data into a hand rehabilitation assessment model pre-trained based on a deep learning algorithm to obtain a hand rehabilitation evaluation result, and then adjust the personalized hand rehabilitation exercise program for the stroke patient according to the hand rehabilitation evaluation result, and then transmit a voice signal for introducing the personalized hand rehabilitation exercise program to the voice speaker, and finally extract the voice instruction of the stroke patient from the voice signal from the sound pickup module, and control the glove body to execute the personalized hand rehabilitation exercise program according to the voice instruction.

[0012] Based on the above invention, a new hand function recovery solution for promoting hand rehabilitation of stroke patients based on generative artificial intelligence technology and voice interaction technology is provided, which includes a glove body for wearing on the hand of a stroke patient and assisting the hand in rehabilitation exercise, and a control module, a data acquisition module, a sound pickup module and a voice speaker configured on the glove body. The control module is used to extract feature data from the hand motion state data received in real time during the completion process and the individual basic information data and / or individual historical exercise data of the stroke patient after completing at least one set of hand rehabilitation exercise movements each time, and then import the feature data into a deep learning-based system. The learning algorithm completes the pre-trained hand rehabilitation assessment model and obtains the hand rehabilitation evaluation results. Then, the personalized hand rehabilitation exercise plan for stroke patients is adjusted according to the hand rehabilitation evaluation results. Then, a voice signal for introducing the exercise plan is transmitted to the voice speaker. Finally, the voice command of the stroke patient is extracted from the voice signal from the sound pickup module, and the glove body is controlled to execute the exercise plan according to the voice command. In this way, through innovation and combination in intelligence, interactivity, personalization and data utilization, it can meet the patient's efficient, convenient and personalized rehabilitation needs, help patients improve their hand motor ability, improve their quality of life and rehabilitation training effects, and facilitate practical application and promotion.

[0013] In one possible design, the data acquisition module includes multiple inertial measurement units, multiple posture sensors, multiple acceleration sensors, multiple angle sensors, multiple pressure sensors, and / or multiple muscle electrical sensors, which are respectively arranged in a one-to-one correspondence at multiple key parts of the glove, wherein the multiple key parts of the glove include fingers, palms, and / or wrists;

[0014] And / or, the individual basic information data contains the patient's age, patient's gender, patient's disease type and / or patient's hand injury degree;

[0015] And / or, the individual historical exercise data contains the historical exercise duration, historical exercise frequency and / or historical exercise effects of each group of hand rehabilitation exercise movements.

[0016] In one possible design, the neural network structure of the deep learning algorithm adopts a recurrent neural network structure, a variant structure of a recurrent neural network, or a combination of a convolutional neural network and a recurrent neural network.

[0017] In a possible design, the hand rehabilitation assessment results include an assessment result of the current rehabilitation stage of the stroke patient, an assessment result of the current recovery degree of various hand functions, and / or a prediction result of rehabilitation progress in the future period.

[0018] In one possible design, the control module is also used to transmit a voice signal for giving rehabilitation encouragement to the voice speaker when it is determined that a preset encouragement trigger condition is met, transmit a voice signal for giving rehabilitation reminder to the voice speaker when it is determined that a preset reminder trigger condition is met, and / or transmit a voice signal for giving rehabilitation advice to the voice speaker when it is determined that a preset suggestion trigger condition is met, based on the hand movement state data received in real time during this completion process and the individual basic information data and / or individual historical exercise data of the stroke patient.

[0019] In one possible design, the glove body includes a wrist portion, a palm portion, and five finger portions, wherein the back side of the finger portion has a full length, and the inner side of the finger portion has a half-finger length and retains a finger cover portion;

[0020] A pneumatic telescopic tube is provided on the back side of each finger portion, and each pneumatic telescopic tube is connected to a manual inflatable piece or an electric air pump through an air valve, wherein the controlled end of the electric air pump is communicatively connected to the control module.

[0021] In one possible design, the device further includes a hedgehog-shaped grip ball, a first elastic rope, and a second elastic rope. The middle portion of the first elastic rope passes through one end of the diameter of the hedgehog-shaped grip ball, forming the entire rope in a double-stranded form. The middle portion of the second elastic rope passes through the other end of the diameter of the hedgehog-shaped grip ball, forming the entire rope in a double-stranded form.

[0022] The two ends of the first elastic rope are respectively connected to one side of the glove body, and the two ends of the second elastic rope are respectively connected to the other side of the glove body, so that the hedgehog-shaped grip ball is closely attached to the palm of the glove body.

[0023] In one possible design, a carbon fiber heating sheet and a temperature sensor are provided in the palm interlayer and / or the back interlayer of the glove body. The glove body is further configured with a power module and a heating drive module. The power module, the heating drive module, and the carbon fiber heating sheet are electrically connected in sequence, and the output end of the temperature sensor and the controlled end of the heating drive module are respectively communicatively connected to the control module.

[0024] The control module is further used to control the driving current output by the heating drive module to the carbon fiber heating plate using PID temperature control technology based on the temperature data collected from the temperature sensor, so as to control the heating of the gloves within the target temperature range of 30 to 50 degrees Celsius.

[0025] In one possible design, the glove body is made of a smart fiber with a "non-von Neumann architecture," wherein the smart fiber has a three-layer sheath-core structure: a core layer is a fiber antenna that induces an alternating electromagnetic field, a middle layer is a dielectric layer for improving the electromagnetic energy coupling capacity, and an outer layer is a luminous layer that is sensitive to electric fields.

[0026] In one possible design, the wrist and / or palm of the glove body is provided with Velcro for allowing the tightness of the glove to be adjusted according to the size of the hand.

[0027] Beneficial effects of the above scheme:

[0028] (1) The present invention provides a novel hand function recovery solution for promoting hand rehabilitation of stroke patients based on generative artificial intelligence technology and voice interaction technology, namely, a glove body for wearing on the hand of a stroke patient and assisting the hand in rehabilitation exercise, and a control module, a data acquisition module, a sound pickup module and a voice speaker configured on the glove body, wherein the control module is used to extract feature data from the hand motion state data received in real time during the completion process and the individual basic information data and / or individual historical exercise data of the stroke patient after completing at least one set of hand rehabilitation exercise movements each time, and then import the feature data into the data acquisition module. A pre-trained hand rehabilitation assessment model is completed based on a deep learning algorithm to obtain hand rehabilitation assessment results. The personalized hand rehabilitation exercise program for stroke patients is then adjusted based on the hand rehabilitation assessment results. A voice signal introducing the exercise program is then transmitted to the voice speaker. Finally, the stroke patient's voice instructions are extracted from the voice signal from the sound pickup module, and the glove body is controlled to execute the exercise program based on the voice instructions. This innovation and combination of intelligence, interactivity, personalization, and data utilization can meet patients' needs for efficient, convenient, and personalized rehabilitation, helping them improve their hand motor skills, quality of life, and rehabilitation training effectiveness.

[0029] (2) Existing devices mainly consider neurological rehabilitation, mechanical fixation, and forced movement, while this solution can focus on promoting circulation, flexible fixation, and increasing interactivity;

[0030] (3) Existing glove designs cannot meet the needs of all users in terms of size. The gloves of this solution have adjustable wrist and finger parts to adapt to different hand sizes and rehabilitation stages. Velcro is used on the wrist and palm to allow users to adjust the tightness of the gloves according to the size of their hands. The dorsal side of the fingers is full length and the inner side is half finger length to ensure the applicability of fingers of different lengths while ensuring the realization of flexion and extension movements.

[0031] (4) The low durability and poor breathability of existing glove materials will lead to poor patient recovery experience and compliance. However, the glove material of this solution uses a new type of smart fiber with a "non-Neumann architecture", which can meet people's needs for textile functionality and comfort while also taking into account the beautiful visual experience;

[0032] (5) The temperature range of gloves with heating devices is currently not limited, which may cause burns, and the heater may not ensure that all parts of the hand receive sufficient heat. However, the gloves of this solution use flexible carbon fiber heating sheets integrated in the palm and back interlayers to achieve heating, with a wide and uniform heating range. The heating range is set at 30-50°C, which is designed to be wide based on the human body's comfortable temperature to adapt to different needs and individual differences. In addition, the patient can adjust the temperature of the gloves themselves through convenient voice interaction, thereby reducing the pressure on caregivers.

[0033] (6) Existing devices that combine grip balls generally use ordinary smooth grip balls, which have limited pressing effects on muscles and are prone to slipping. However, this solution uses a hedgehog-shaped grip ball, which is more conducive to pressure stimulation; and uses two elastic ropes to pass through two protrusions at both ends of the diameter of the grip ball, and sew them to both sides of the glove in a double-strand rope, so as to achieve the effect of firmly attaching the hedgehog-shaped grip ball to the palm of the glove;

[0034] (7) The current device that uses elastic drawstrings to achieve finger bending has unstable force maintenance, and the gloves that use straightening and resetting devices to achieve finger straightening effects are prone to causing additional pressure on the patient's fingers and causing discomfort. However, the gloves of this scheme use a telescopic tube pneumatic device to help the hands establish correct movement patterns and improve hand muscle strength through repeated finger flexion and extension movements. In the process, all parts of the fingers can be bent arbitrarily and have good elasticity, and the training is gentle;

[0035] (8) This solution can take into account the needs of different groups of people. In addition to using the electric pump for automatic inflation, it also considers that some patients want to buy cheaper products, and uses manual ball inflation as a supplement, reflecting universal benefits;

[0036] (9) This program greatly improves the pertinence and effectiveness of rehabilitation training through intelligent rehabilitation training program adjustment and personalized services. It can customize training according to the unique situation of each patient, significantly improving the rehabilitation effect and shortening the patient's rehabilitation cycle;

[0037] (10) The voice interaction function of this solution greatly enhances the patients’ sense of participation and enthusiasm in the rehabilitation process. That is, patients are no longer passively receiving training, but can actively communicate with and control the equipment, which increases their willingness to persist in rehabilitation training, thus helping to better restore hand function;

[0038] (11) The precise data collection and analysis of this program can enable rehabilitation therapists and patients to understand the rehabilitation progress more clearly, identify problems in a timely manner and adjust the training plan, thus avoiding blind training and making the rehabilitation process more scientific and reasonable;

[0039] (12) The comfortable material and ergonomic design of this solution not only ensure the patient's comfort during use;

[0040] (13) This solution reduces the resistance caused by discomfort and increases the frequency and duration of use by patients;

[0041] (14) The innovative rehabilitation method of this program reduces the workload of physical therapists, improves the utilization efficiency of medical resources, provides high-quality rehabilitation services to more patients, and facilitates practical application and promotion. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0043] Figure 1 A schematic structural diagram of a glove device for promoting hand rehabilitation of stroke patients based on intelligent voice interaction provided in an embodiment of the present invention.

[0044] Figure 2 A schematic diagram of the wearing structure of the glove device provided by an embodiment of the present invention.

[0045] Figure 3 This is a schematic structural diagram of the back of the hand side of the glove body in the glove device provided by an embodiment of the present invention.

[0046] Figure 4 Schematic diagram of the communication structure of the pneumatic telescopic tube, air valve and manual inflator in the glove device provided by an embodiment of the present invention.

[0047] Figure 5 This is a schematic diagram of the combined structure of the palm portion of the glove body and the hedgehog-shaped grip ball in the glove device provided by an embodiment of the present invention.

[0048] Figure 6 This is a schematic diagram of the palm side structure of the glove body in the glove device provided by an embodiment of the present invention.

[0049] In the above drawings: 1-glove body; 14-pneumatic telescopic tube; 15-air valve; 16-electric air pump; 17-inflatable ball; 18-velcro; 19-reinforced Velcro; 5-voice speaker; 6-hedgehog-shaped grip ball; 71-first elastic rope; 72-second elastic rope. DETAILED DESCRIPTION

[0050] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the present invention will be briefly introduced below in conjunction with the drawings and the description of the embodiments or the prior art. Obviously, the following description of the structures of the drawings is only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained based on these embodiments without creative work. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention.

[0051] It should be understood that although the terms first, second, etc. may be used herein to describe various objects, these objects should not be limited by these terms. These terms are merely used to distinguish one object from another. For example, a first object can be referred to as a second object, and similarly, a second object can be referred to as a first object without departing from the scope of the exemplary embodiments of the present invention.

[0052] It should be understood that the term "and / or" that may appear in this document is merely a description of the association relationship between associated objects, indicating that there may be three relationships. For example, A and / or B can indicate three situations: A exists alone, B exists alone, or A and B exist at the same time. For another example, A, B and / or C can indicate the existence of any one of A, B and C or any combination of them. The term " / and" that may appear in this document describes another type of association object relationship, indicating that there may be two relationships. For example, A / and B can indicate two situations: A exists alone or A and B exist at the same time. In addition, the character " / " that may appear in this document generally indicates that the previous and next associated objects are in an "or" relationship.

[0053] Example

[0054] like Figures 1 to 6 As shown, the glove device provided in this embodiment promotes hand rehabilitation of stroke patients based on intelligent voice interaction, including but not limited to a glove body 1 for wearing on the hand of a stroke patient and assisting the hand in rehabilitation exercises, and a control module, a data acquisition module, a sound pickup module and a voice speaker 5 configured on the glove body 1.

[0055] The data acquisition module is communicatively connected to the control module and is used to collect hand motion state data of the stroke patient in real time and transmit the hand motion state data to the control module in real time. The hand motion state data is used to reflect the patient's actual hand movements during rehabilitation training or daily activities. Specifically, the data acquisition module includes, but is not limited to, multiple inertial measurement units, multiple attitude sensors, multiple acceleration sensors, multiple angle sensors, multiple muscle electrical sensors, and / or multiple pressure sensors, which are arranged in a one-to-one correspondence at multiple key locations of the glove. The multiple key locations of the glove include, but are not limited to, the fingers, palm, and / or wrist. Through the aforementioned data acquisition module, hand motion state data including, but not limited to, angle changes of various hand joints, acceleration, angular velocity, muscle electrical signals, motion trajectory, and the force applied to the hand gripping movement can be collected. In addition, the aforementioned inertial measurement units, attitude sensors, acceleration sensors, angle sensors, pressure sensors, and muscle electrical sensors can all be implemented using existing related products.

[0056] The sound pickup module is communicatively connected to the control module and is used to collect voice signals from the stroke patient in real time and transmit the voice signals to the control module in real time. Specifically, the sound pickup module can be implemented by, but is not limited to, a microphone product.

[0057] The voice speaker 5 is communicatively connected to the control module and is used to play the voice signal from the control module.

[0058] The control module is used to extract feature data from the hand motion state data received in real time during the completion process and the individual basic information data and / or individual historical exercise data of the stroke patient after completing at least one set of hand rehabilitation exercise movements each time, and then import the feature data into a hand rehabilitation assessment model pre-trained based on a deep learning algorithm to obtain a hand rehabilitation evaluation result, and then adjust the personalized hand rehabilitation exercise program for the stroke patient based on the hand rehabilitation evaluation result, and then transmit a voice signal for introducing the personalized hand rehabilitation exercise program to the voice speaker 5, and finally extract the voice command of the stroke patient from the voice signal from the sound pickup module, and control the glove body 1 to execute the personalized hand rehabilitation exercise program according to the voice command. The hand rehabilitation exercise movement refers to the conventional hand movements such as fist clenching, finger extension and wrist rotation performed by the patient after wearing the glove body 1 according to some set standards. Considering that patients of different age groups and genders may have different rehabilitation abilities and speeds, and patients with different types of injuries may have different rehabilitation processes and priorities, specifically, the individual basic information data includes but is not limited to the patient's age, gender, type of patient condition, and / or degree of hand injury, which can be obtained through conventional input by the patient or rehabilitation therapist. In addition, the individual historical exercise data specifically includes but is not limited to the historical exercise duration, historical exercise frequency, and / or historical exercise effects of each group of hand rehabilitation exercise movements, in order to help the model better understand the patient's rehabilitation trajectory and trends, which can be obtained based on the analysis of the historically collected hand movement state data.

[0059] The feature data extracted from the hand movement state data received in real time during this completion process and the individual basic information data and / or individual historical exercise data of the stroke patient specifically include but are not limited to: performing conventional denoising on the collected hand movement state data to remove abnormal data points caused by environmental interference, sensor errors, and other factors to ensure the accuracy and stability of the data; performing standardization on the denoised data to unify data of different dimensions (for example, data collected by different sensors with large differences in range) into a reasonable scale range to facilitate subsequent analysis and processing; based on the hand movement state data, using time domain analysis methods to extract feature data of hand movement from the perspective of time series, such as feature data such as the duration and frequency of the movement; based on the hand movement state data, using frequency domain analysis methods (for example, through fast Fourier transform, etc.) to analyze features such as energy distribution of hand movement at different frequencies, which can reflect the stability and coordination of hand movement; extracting key semantic information from the individual basic information data, such as extracting the criticality description of the rehabilitation goal, key hand function features, and so on. Finally, all the extracted feature data are integrated to form a complete set of feature data that can reflect the comprehensive condition of the patient's hand.

[0060] The deep learning algorithm (Deep Learning) specifically refers to a machine learning algorithm based on deep neural network models and methods; it is developed on the basis of algorithm models such as statistical machine learning and artificial neural networks, combined with the development of contemporary big data and large computing power. The most important technical feature of deep learning is the ability to automatically extract features. The extracted features are also called deep features or deep feature representations. Compared with manually designed features, deep features have stronger and more robust representation capabilities. Specifically, the neural network structure of the deep learning algorithm can be, but is not limited to, a recurrent neural network structure, a variant structure of a recurrent neural network (such as a long short-term memory network LSTM or a gated recurrent unit GRU, etc.) or a combination of a convolutional neural network and a recurrent neural network (wherein the convolutional neural network first extracts local spatial features, and the recurrent neural network then processes time series features, which is suitable for more complex hand movement analysis scenarios); since hand movement is a process with time series characteristics, the use of the aforementioned model structure can have a great advantage in processing sequence data and can well capture the associated information before and after hand movement. Based on a certain amount of sample data (the model input is the feature data, and the model output is the hand rehabilitation assessment result label), a validated hand rehabilitation assessment model can be trained using conventional calibration and verification modeling methods (the specific process includes model calibration and verification, that is, first comparing the model simulation results with the measured data, and then adjusting the model parameters based on the comparison results to ensure that the simulation results match the actual data). During the calibration and verification modeling process of the hand rehabilitation assessment model, a tree-structured Bayesian optimization algorithm can be used to tune the model parameters. In addition, the hand rehabilitation assessment results specifically include but are not limited to the assessment results of the current rehabilitation stage of the stroke patient, the current recovery degree assessment results of various hand functions and / or the rehabilitation progress prediction results in the future period, etc., wherein the current rehabilitation stage assessment result refers to the assessment of the patient's current rehabilitation stage, such as the acute stage, subacute stage or recovery stage, etc., to provide a basis for subsequent adjustments to the rehabilitation training program; the current recovery degree assessment result refers to the recovery degree of various hand functions of the patient, such as finger flexibility, grip strength or hand coordination, etc., expressed in the form of specific numerical values or levels, so that patients and rehabilitation doctors can intuitively understand the rehabilitation effect; the rehabilitation progress prediction result refers to the prediction of the patient's rehabilitation progress in the future period, including the expected rehabilitation stage, further improvement of hand function, etc., to help rehabilitation doctors formulate more forward-looking rehabilitation plans.

[0061] Since the hand rehabilitation evaluation results specifically include but are not limited to the evaluation results of the current rehabilitation stage of the stroke patient, the current recovery degree evaluation results of various hand functions and / or the predicted results of rehabilitation progress in the future, etc., it is possible to combine previous rehabilitation knowledge and the best training strategies corresponding to different movement states, and routinely adjust the personalized hand rehabilitation exercise program of the stroke patient according to the hand rehabilitation evaluation results. The content of the program can include but is not limited to specific training movements, the number of repetitions of each movement, the time interval of training, and the expected stage-by-stage rehabilitation goals. According to the voice command, the glove body 1 is controlled to execute the personalized hand rehabilitation exercise program, including but not limited to: according to the patient's instructions such as "start training", "increase intensity" or "pause", the personalized hand rehabilitation exercise program is started, the personalized hand rehabilitation exercise program is strengthened, or the personalized hand rehabilitation exercise program is paused, etc. In addition, the control module can be specifically but not limited to being implemented by a microcontroller module chip of the STM32 series.

[0062] Based on the detailed description of the glove device described above, it can be seen that this embodiment provides a new hand function recovery solution for promoting hand rehabilitation of stroke patients based on generative artificial intelligence technology and voice interaction technology, that is, it includes a glove body for wearing on the hand of a stroke patient and assisting the hand in rehabilitation exercise, and a control module, a data acquisition module, a sound pickup module and a voice speaker configured on the glove body, wherein the control module is used to extract feature data from the hand movement state data received in real time during the completion process and the individual basic information data and / or individual historical exercise data of the stroke patient after completing at least one set of hand rehabilitation exercise movements each time, and then import the feature data into the glove body. A pre-trained hand rehabilitation assessment model based on a deep learning algorithm is input to obtain the hand rehabilitation evaluation results. Then, the personalized hand rehabilitation exercise plan for stroke patients is adjusted according to the hand rehabilitation evaluation results. Then, a voice signal for introducing the exercise plan is transmitted to the voice speaker. Finally, the voice command of the stroke patient is extracted from the voice signal from the sound pickup module, and the glove body is controlled to execute the exercise plan according to the voice command. In this way, through innovation and combination in intelligence, interactivity, personalization and data utilization, it can meet the patient's efficient, convenient and personalized rehabilitation needs, help patients improve their hand motor ability, improve their quality of life and rehabilitation training effects, and facilitate practical application and promotion.

[0063] Preferably, the control module is also used to transmit a voice signal for giving rehabilitation encouragement to the voice speaker 5 when it is determined that the preset encouragement trigger condition is met, transmit a voice signal for giving rehabilitation reminder to the voice speaker 5 when it is determined that the preset reminder trigger condition is met, and / or transmit a voice signal for giving rehabilitation advice to the voice speaker 5 when it is determined that the preset suggestion trigger condition is met, based on the hand movement state data received in real time during this completion process and the individual basic information data and / or individual historical exercise data of the stroke patient. The encouragement triggering conditions can be specifically but not limited to the following three aspects: (1) Action completion, that is, when the patient successfully completes a set of rehabilitation training actions and the completion quality of the actions reaches or exceeds the preset standard, the encouragement voice is triggered; for example, if the patient completes 10 fist-clenching actions as required, and the strength of each fist-clenching and the degree of finger extension meet the requirements, encouragement will be given; (2) Significant progress, that is, by comparing the patient's current hand movement data with previous historical data, if it is found that the patient has made significant progress in a certain action or overall hand function, such as a significant improvement in hand flexibility, coordination or strength, encouragement will be given in a timely manner; (3) Persistence in training, that is, when the patient continuously performs rehabilitation training for a certain time or number of times, even if the action completion situation does not change significantly, but the patient's persistence and efforts are worthy of recognition, the encouragement voice will be triggered. The reminder triggering conditions can be specifically but not limited to the following three aspects: (4) Action standardization, that is, when the patient is performing rehabilitation exercises, if it is found that his actions are not standard, a reminder will be given in a timely manner. For example, reminding patients to “stretch your fingers as straight as possible and not bend them” or “use even force when making a fist and do not rely solely on the thumb”, etc., to help patients complete the movements correctly and achieve better rehabilitation effects; (5) Training progress, that is, informing patients of the current training progress, such as “you have completed half of today’s training plan, keep going” or “this is the third set of movements, and there are two more to go”, etc., so that patients have a clear understanding of the training process and enhance their confidence and motivation to complete the training; (6) Precautions, that is, reminding patients of precautions during rehabilitation training, such as “if you feel pain or discomfort during training, stop immediately” or “keep breathing smooth and do not hold your breath”, etc., to ensure the safety of patients’ training.The triggering conditions for the suggestions can be specifically, but not limited to, triggered from the following three aspects: (7) Based on model analysis, that is, based on the patient's real-time hand movement status and rehabilitation needs, combined with existing rehabilitation training data and professional knowledge, generate personalized rehabilitation suggestions; for example, if the model analysis finds that the patient's wrist flexibility is poor, the patient may be advised to add some wrist rotation and flexion and extension training movements; (8) Combined with the rehabilitation stage, that is, according to the patient's rehabilitation stage, corresponding rehabilitation suggestions will be provided; in the early stage of rehabilitation, the patient may be advised to perform some simple hand extension and fisting movements to relieve hand stiffness; in the middle stage of rehabilitation, the difficulty and intensity of the movements will be gradually increased, such as using a hand gripper for training; in the late stage of rehabilitation, the patient will be advised to perform some functional training, such as simulating movements in daily life to improve the practicality of the hand; (9) Reference to professional knowledge, that is, professional rehabilitation medicine knowledge and clinical experience related content searched on the Internet will be referred to provide suggestions, making the rehabilitation suggestions more scientific and professional. Since the encouragement, prompts and rehabilitation suggestions given to patients through voice can enhance the patient's sense of participation and enthusiasm in the rehabilitation process.

[0064] Preferably, the glove body 1 includes but is not limited to a wrist portion, a palm portion and five finger portions, wherein the back side of the finger portion has a full length, and the inner side of the finger portion has a half finger length and retains a finger cover portion; a pneumatic telescopic tube 14 is provided on the back side of each finger portion, and each pneumatic telescopic tube 14 is connected to a manual inflatable part or an electric air pump 16 through an air valve 15, wherein the controlled end of the electric air pump 16 is communicatively connected to the control module. The aforementioned design of the back inner side length of the finger portion can ensure the applicability of fingers of different lengths, while ensuring the realization of flexion and extension movements. Figures 3-4 As shown, based on the design of the aforementioned pneumatic telescopic tube 14 and air valve 15, by inflating the telescopic tube, it gradually extends and drives the fingers to bend and clench into a fist. After deflating, the telescopic tube returns to its original shape, and the fingers also stretch. Repeated finger flexion and extension exercises help the hand establish a correct movement pattern, improve hand muscle strength, and facilitate the rehabilitation of hand function in stroke patients. In addition, based on the inflation and deflation structure design of the aforementioned manual inflatable member or electric inflatable pump 16, the inflation method can be manual or automatic. Among them, the manual inflatable member can be implemented using the inflatable ball 17, which is compact and lightweight, easy to carry, low cost, and can still be used even without power. The automatic method uses an electric pump, which is easy and fast, saves time and effort, can accurately control the air pressure, and can be inflated and deflated at regular intervals. The aforementioned two modes take into account the user's family economic situation. Families who think that electric inflatable pumps are too expensive can choose manual inflation and deflation.

[0065] Preferably, it also includes but is not limited to a hedgehog-shaped grip ball 6, a first elastic cord 71, and a second elastic cord 72, wherein the middle portion of the first elastic cord 71 passes through one end of the diameter of the hedgehog-shaped grip ball 6 and makes the entire cord body a double-strand rope, and the middle portion of the second elastic cord 72 passes through the other end of the diameter of the hedgehog-shaped grip ball 6 and makes the entire cord body also a double-strand rope; the two ends of the first elastic cord 71 are respectively connected to one side of the glove body 1, and the two ends of the second elastic cord 72 are respectively connected to the other side of the glove body 1, and the hedgehog-shaped grip ball 6 is pressed against the palm of the glove body 1. Figure 5 As shown, by configuring the hedgehog-shaped grip ball 6, the soft spines on the ball can stimulate the nerves and blood vessels in the hand during flexion and extension exercises, thereby promoting blood circulation and helping to relieve numbness and stiffness in the hand. Furthermore, the design of the first elastic cord 71 and the second elastic cord 72 can enhance the stability of the hedgehog-shaped grip ball 6 in the palm of the glove body 1, ensuring its effectiveness. Furthermore, the properties of the hedgehog-shaped grip ball 6 may include, but are not limited to: a diameter of 7 cm, a weight of 15 pounds, and being made of rubber.

[0066] Preferably, a carbon fiber heating sheet and a temperature sensor are provided in the palm interlayer and / or back interlayer of the glove body 1. The glove body 1 is also equipped with a power module and a heating drive module, wherein the power module, the heating drive module, and the carbon fiber heating sheet are electrically connected in sequence, and the output end of the temperature sensor and the controlled end of the heating drive module are respectively communicatively connected to the control module. The control module is further configured to control the driving current output by the heating drive module to the carbon fiber heating sheet based on the temperature data collected from the temperature sensor using PID temperature control technology, so as to control the heating of the glove within the target temperature range of 30 to 50 degrees Celsius. The power module can also power the control module, data acquisition module, sound pickup module, and voice speaker 5. Specifically, it can be composed of two 3V button batteries connected in series to achieve the purpose of small size and lightness. It can also be in the form of an auxiliary battery (such as a charger) that converts electromagnetic energy to reduce energy consumption. The heating drive module can specifically use a relay (e.g., a relay model SRD-03VDC-SL-C) to control the on / off switching of the driving current to achieve current regulation (even if the driving current is a PWM signal, the driving current can be adjusted by adjusting the on / off frequency). The PID temperature control technology achieves precise temperature control by adjusting three parameters: proportional (P), integral (I), and differential (D). Through the aforementioned heating design, a wide and uniform heating range can be achieved, and the heating range is set between 30 and 50°C. This also allows for a wide design based on the human body's comfortable temperature to accommodate different needs and individual differences. In addition, the control module can also control the driving current output by the heating drive module to the carbon fiber heating plate based on the voice command (e.g., key commands such as "increase temperature" or "lower temperature"), so that the patient can adjust the temperature of the glove themselves through convenient voice interaction, thereby reducing the pressure on caregivers.

[0067] Preferably, the material of the glove body 1 adopts a smart fiber with a "non-von Neumann architecture", wherein the smart fiber has the following three-layer sheath-core structure: the core layer is a fiber antenna that induces an alternating electromagnetic field, the middle layer is a dielectric layer for improving the electromagnetic energy coupling capacity, and the outer layer is a luminous layer that is sensitive to the electric field. The smart fiber can specifically be, but is not limited to, a new type of "non-von Neumann architecture" smart fiber newly developed by Donghua University. The fiber has a three-layer sheath-core structure, and all materials used are relatively common on the market: the core layer is a fiber antenna that induces an alternating electromagnetic field (for example, silver-plated nylon fiber), the middle layer is a dielectric layer for improving the electromagnetic energy coupling capacity (for example, BaTiO3 composite resin), and the outer layer is an electric field-sensitive luminescent layer (for example, ZnS composite resin). The glove body 1 woven from the smart fiber has good durability, good comfort, good breathability, good skin-friendliness, light weight, and is non-allergenic. It can also emit blue light upon contact with human skin, and has a cool and beautiful appearance. While meeting people's demand for functionality and comfort in textiles, it also takes into account the aesthetic visual experience, solving the problems existing in previous materials.

[0068] Preferably, the wrist and / or palm of the glove body 1 is provided with a Velcro 18 for adjusting the tightness of the glove according to the size of the hand. Figure 6 As shown, specifically, the palm portion is specifically provided with a Velcro 18 for allowing the tightness of the glove to be adjusted according to the size of the hand, and a reinforcing Velcro 19 for reinforcing the Velcro 18. Through the above design, the wearability can be further improved.

[0069] In summary, the glove device provided in this embodiment has the following technical effects:

[0070] (1) This embodiment provides a new hand function recovery solution for promoting hand rehabilitation of stroke patients based on generative artificial intelligence technology and voice interaction technology, namely, a glove body for wearing on the hand of a stroke patient and assisting the hand in rehabilitation exercise, and a control module, a data acquisition module, a sound pickup module and a voice speaker configured on the glove body, wherein the control module is used to extract feature data from the hand movement state data received in real time during the completion process and the individual basic information data and / or individual historical exercise data of the stroke patient after completing at least one set of hand rehabilitation exercise movements each time, and then import the feature data into the base A pre-trained hand rehabilitation assessment model is completed using a deep learning algorithm to obtain hand rehabilitation assessment results. Based on these results, the personalized hand rehabilitation exercise program for stroke patients is adjusted. A voice signal introducing the exercise program is then transmitted to the voice speaker. Finally, the voice signal from the sound pickup module is extracted to obtain the stroke patient's voice instructions. The glove itself is then controlled to execute the exercise program based on the voice instructions. Through innovation and integration in intelligence, interactivity, personalization, and data utilization, this system can meet patients' needs for efficient, convenient, and personalized rehabilitation, helping them improve their hand motor skills, quality of life, and rehabilitation training effectiveness.

[0071] (2) Existing devices mainly consider neurological rehabilitation, mechanical fixation, and forced movement, while this solution can focus on promoting circulation, flexible fixation, and increasing interactivity;

[0072] (3) Existing glove designs cannot meet the needs of all users in terms of size. The gloves of this solution have adjustable wrist and finger parts to adapt to different hand sizes and rehabilitation stages. Velcro is used on the wrist and palm to allow users to adjust the tightness of the gloves according to the size of their hands. The dorsal side of the fingers is full length and the inner side is half finger length to ensure the applicability of fingers of different lengths while ensuring the realization of flexion and extension movements.

[0073] (4) The low durability and poor breathability of existing glove materials will lead to poor patient recovery experience and compliance. However, the glove material of this solution uses a new type of smart fiber with a "non-Neumann architecture", which can meet people's needs for textile functionality and comfort while also taking into account the beautiful visual experience;

[0074] (5) The temperature range of gloves with heating devices is currently not limited, which may cause burns, and the heater may not ensure that all parts of the hand receive sufficient heat. However, the gloves of this solution use flexible carbon fiber heating sheets integrated in the palm and back interlayers to achieve heating, with a wide and uniform heating range. The heating range is set at 30-50°C, which is designed to be wide based on the human body's comfortable temperature to adapt to different needs and individual differences. In addition, the patient can adjust the temperature of the gloves themselves through convenient voice interaction, thereby reducing the pressure on caregivers.

[0075] (6) Existing devices that combine grip balls generally use ordinary smooth grip balls, which have limited pressing effects on muscles and are prone to slipping. However, this solution uses a hedgehog-shaped grip ball, which is more conducive to pressure stimulation; and uses two elastic ropes to pass through two protrusions at both ends of the diameter of the grip ball, and sew them to both sides of the glove in a double-strand rope, so as to achieve the effect of firmly attaching the hedgehog-shaped grip ball to the palm of the glove;

[0076] (7) The current device that uses elastic drawstrings to achieve finger bending has unstable force maintenance, and the gloves that use straightening and resetting devices to achieve finger straightening effects are prone to causing additional pressure on the patient's fingers and causing discomfort. However, the gloves of this scheme use a telescopic tube pneumatic device to help the hands establish correct movement patterns and improve hand muscle strength through repeated finger flexion and extension movements. In the process, all parts of the fingers can be bent arbitrarily and have good elasticity, and the training is gentle;

[0077] (8) This solution can take into account the needs of different groups of people. In addition to using the electric pump for automatic inflation, it also considers that some patients want to buy cheaper products, and uses manual ball inflation as a supplement, reflecting universal benefits;

[0078] (9) This program greatly improves the pertinence and effectiveness of rehabilitation training through intelligent rehabilitation training program adjustment and personalized services. It can customize training according to the unique situation of each patient, significantly improving the rehabilitation effect and shortening the patient's rehabilitation cycle;

[0079] (10) The voice interaction function of this solution greatly enhances the patients’ sense of participation and enthusiasm in the rehabilitation process. That is, patients are no longer passively receiving training, but can actively communicate with and control the equipment, which increases their willingness to persist in rehabilitation training, thus helping to better restore hand function;

[0080] (11) The precise data collection and analysis of this program can enable rehabilitation therapists and patients to understand the rehabilitation progress more clearly, identify problems in a timely manner and adjust the training plan, thus avoiding blind training and making the rehabilitation process more scientific and reasonable;

[0081] (12) The comfortable material and ergonomic design of this solution not only ensure the patient's comfort during use;

[0082] (13) This solution reduces the resistance caused by discomfort and increases the frequency and duration of use by patients;

[0083] (14) The innovative rehabilitation method of this program reduces the workload of physical therapists, improves the utilization efficiency of medical resources, provides high-quality rehabilitation services to more patients, and facilitates practical application and promotion.

[0084] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included in the scope of protection of the present invention.

Claims

1. A glove device based on intelligent voice interaction to promote hand rehabilitation of stroke patients, characterized in that: The invention comprises a glove body (1) for being worn on the hand of a stroke patient and assisting the hand in rehabilitation exercises, and a control module, a data acquisition module, a sound pickup module and a voice speaker (5) arranged on the glove body (1); The data acquisition module is communicatively connected to the control module, and is used to collect the hand movement state data of the stroke patient in real time, and transmit the hand movement state data to the control module in real time; The sound pickup module is communicatively connected to the control module and is used to collect voice signals from the stroke patient in real time and transmit the voice signals to the control module in real time; The voice speaker (5) is communicatively connected to the control module and is used to play the voice signal from the control module; The control module is used to extract feature data from the hand movement state data and the individual basic information data and / or individual historical exercise data of the stroke patient received in real time during the completion process after completing at least one set of hand rehabilitation exercise movements each time, and then import the feature data into a hand rehabilitation evaluation model pre-trained based on a deep learning algorithm to obtain a hand rehabilitation evaluation result, and then adjust the personalized hand rehabilitation exercise program for the stroke patient according to the hand rehabilitation evaluation result, and then transmit a voice signal for introducing the personalized hand rehabilitation exercise program to the voice speaker (5), and finally extract the voice instruction of the stroke patient from the voice signal from the sound pickup module, and control the glove body (1) to execute the personalized hand rehabilitation exercise program according to the voice instruction.

2. The glove device according to claim 1, wherein: The data acquisition module includes a plurality of inertial measurement units, a plurality of posture sensors, a plurality of acceleration sensors, a plurality of angle sensors, a plurality of pressure sensors and / or a plurality of muscle electrical sensors, which are respectively arranged in a one-to-one correspondence at a plurality of key parts of the glove, wherein the plurality of key parts of the glove include the fingers, the palm and / or the wrist; And / or, the individual basic information data contains the patient's age, patient's gender, patient's disease type and / or patient's hand injury degree; And / or, the individual historical exercise data contains the historical exercise duration, historical exercise frequency and / or historical exercise effects of each group of hand rehabilitation exercise movements.

3. The glove device according to claim 1, wherein: The neural network structure of the deep learning algorithm adopts a recurrent neural network structure, a variant structure of a recurrent neural network, or a combination structure of a convolutional neural network and a recurrent neural network.

4. The glove device according to claim 1, wherein: The hand rehabilitation assessment results include the assessment results of the current rehabilitation stage of the stroke patient, the assessment results of the current recovery degree of various hand functions and / or the prediction results of rehabilitation progress in the future period.

5. The glove device according to claim 1, wherein: The control module is also used to transmit a voice signal for giving rehabilitation encouragement to the voice speaker (5) when it is determined that a preset encouragement trigger condition has been met, transmit a voice signal for giving rehabilitation reminder to the voice speaker (5) when it is determined that a preset reminder trigger condition has been met, and / or transmit a voice signal for giving rehabilitation advice to the voice speaker (5) when it is determined that a preset suggestion trigger condition has been met, based on the hand movement state data received in real time during the completion process and the individual basic information data and / or individual historical exercise data of the stroke patient.

6. The glove device according to claim 1, wherein: The glove body (1) comprises a wrist portion, a palm portion and five finger portions, wherein the back side of the finger portion has a full length, and the inner side of the finger portion has a half finger length and retains a finger cover portion; A pneumatic telescopic tube (14) is provided on the back side of each finger portion, and each pneumatic telescopic tube (14) is connected to a manual inflatable component or an electric inflatable pump (16) through an air valve (15), wherein a controlled end of the electric inflatable pump (16) is communicatively connected to the control module.

7. The glove device according to claim 1, wherein: It also includes a hedgehog-shaped grip ball (6), a first elastic rope (71) and a second elastic rope (72), wherein the middle portion of the first elastic rope (71) passes through one end of the diameter of the hedgehog-shaped grip ball (6) and makes the entire rope body in the form of a double rope, and the middle portion of the second elastic rope (72) passes through the other end of the diameter of the hedgehog-shaped grip ball (6) and makes the entire rope body also in the form of a double rope; The two ends of the first elastic rope (71) are respectively connected to one side of the glove body (1), and the two ends of the second elastic rope (72) are respectively connected to the other side of the glove body (1), and the hedgehog-shaped grip ball (6) is made to stick closely to the palm of the glove body (1).

8. The glove device according to claim 1, wherein: A carbon fiber heating sheet and a temperature sensor are provided in the palm interlayer and / or the back interlayer of the glove body (1), and the glove body (1) is further provided with a power supply module and a heating drive module, wherein the power supply module, the heating drive module and the carbon fiber heating sheet are electrically connected in sequence, and the output end of the temperature sensor and the controlled end of the heating drive module are respectively communicatively connected to the control module; The control module is further used to control the driving current output by the heating drive module to the carbon fiber heating plate using PID temperature control technology based on the temperature data collected from the temperature sensor, so as to control the heating of the gloves within the target temperature range of 30 to 50 degrees Celsius.

9. The glove device according to claim 1, wherein: The material of the glove body (1) is an intelligent fiber with a "non-von Neumann architecture", wherein the intelligent fiber has the following three-layer sheath-core structure: the core layer is a fiber antenna for inducing an alternating electromagnetic field, the middle layer is a dielectric layer for improving the electromagnetic energy coupling capacity, and the outer layer is a luminous layer that is sensitive to the electric field.

10. The glove device according to claim 1, wherein: The wrist and / or palm of the glove body (1) is provided with a Velcro (18) for allowing the tightness of the glove to be adjusted according to the size of the hand.

Images