Rehabilitation assessment method and system based on wearable device

Abstract

The invention discloses a rehabilitation treatment evaluation method and system based on wearable equipment, and relates to the technical field of medical rehabilitation. The system comprises a wearable sensing module, a wireless communication module, a cloud data processing module, a man-machine interaction module and an auxiliary stimulation module. The wearable sensing module collects joint movement, muscle movement and stress parameters in a non-invasive mode. According to the method, through closed-loop processes of equipment preprocessing and personalized configuration, adaptive model training, real-time data acquisition, cloud quantitative evaluation and multi-dimensional feedback adjustment, full-chain coverage of'perception-analysis-evaluation-feedback 'of rehabilitation treatment is realized. According to the method, the defect that a cognitive movement therapy lacks quantitative indexes is overcome, the problems that traditional rehabilitation evaluation is subjective and a sensor can only detect and cannot assist in treatment are solved, rehabilitation treatment standardization, individuation and high efficiency are achieved, the labor cost is reduced, the rehabilitation quality is improved, and the method is suitable for rehabilitation scenes such as cerebral apoplexy and movement injury.

Description

Technical Field

[0001] This invention relates to the field of medical rehabilitation technology, and specifically to a rehabilitation treatment assessment method and system based on wearable devices. Background Technology

[0002] With the rise of the Internet of Things, smart homes, and smart healthcare, the demand for posture sensors continues to increase, driving the rapid development of related technologies both domestically and internationally. Domestically, universities, research institutions, and enterprises are actively investing in research and development, continuously improving the accuracy, stability, and reliability of posture sensors, while also perfecting standardization and normalization. For example, the consumer-grade posture sensor launched by Victor Intelligent integrates multiple modules and can reproduce motion through 3D animation. Internationally, related technologies have a longer history of research and application. The University of Waterloo in Canada has achieved significant results in calibration algorithms, and medical sensors from companies such as Bosch Sensortec, Infineon, and Biometrics are widely used in medical devices, health monitoring systems, and other fields, providing reliable health data support. However, currently, sensor technology mostly remains at the assessment and detection level, with limited targeted development for specific rehabilitation treatments (especially exercise therapy), and a lack of standardized rehabilitation assessment methods and a closed-loop system from data acquisition to treatment feedback.

[0003] Cognitive Therapeutic Exercise (CTE), founded by Italian neurologist Carlo Perfetti, is based on cognitive theory and promotes motor function recovery by activating cognitive processes in the brain (perception, attention, memory, judgment, and speech). This therapy emphasizes the interaction between the body and the environment, recognizing the plasticity of the brain. It involves the body transmitting subtle information to the brain, which then filters the necessary information to complete actions. This effectively improves balance and walking function in stroke patients, offering more stable treatment results compared to conventional exercise therapy. It has been widely used both domestically and internationally for post-stroke functional recovery. However, CTE training methods lack quantitative indicators and standardized training guidance. Rehabilitation assessments rely primarily on the subjective observation of medical staff, leading to inconsistent assessment results and difficulty in accurately tracking rehabilitation progress.

[0004] Currently, the increasing maturity of hardware and software technologies, combined with the demands for standardization, proceduralization, and personalization in rehabilitation treatment programs, has created both the possibility and the necessity for interdisciplinary collaboration in rehabilitation engineering. However, existing technologies show a significant gap in their integration with rehabilitation assessment and treatment methods: there is a lack of integrated solutions that deeply combine wearable devices with rehabilitation assessment methods, making it impossible to automate the entire process from data collection and quantitative assessment to treatment feedback. This results in low efficiency, high labor costs, and highly subjective assessment results in rehabilitation treatment. Therefore, a standardized rehabilitation assessment method and a corresponding integrated system are urgently needed to address these technical pain points. Summary of the Invention

[0005] The purpose of this invention is to provide a rehabilitation treatment assessment method and system based on wearable devices to solve the problem of the lack of quantitative indicators in the prior art.

[0006] The technical solution of the present invention to solve the above-mentioned technical problems is as follows: A wearable device-based rehabilitation assessment system consists of five functional modules, which work together to form a complete closed loop for rehabilitation assessment and treatment. The core components of the wearable sensing module include posture sensors, surface electromyography (EMG) sensors, and tension sensors. Each component can be a commercially available mature module or a custom-developed integrated module. The selection must meet the accuracy and stability requirements of the rehabilitation scenario. Attitude sensors: Symmetrically distributed at the midpoint of the outer side of the three main joints of the human body, namely the hip, knee and ankle, integrating gyroscopes, accelerometers, etc., to collect parameters such as relative joint position, angular velocity, and inertial acceleration, with a sampling rate ≥100Hz to ensure the capture of subtle motion changes; Surface electromyography (EMG) sensor: Attached to major muscle groups such as the quadriceps femoris, gastrocnemius, peroneus longus, and biceps femoris, it is used to collect muscle electrical signals during exercise to reflect muscle exertion performance. The sampling rate is ≥200Hz and supports anti-motion artifact processing. Tension sensor: Located behind the popliteal fossa, it is a long strip that fits snugly against the skin. It is used to collect force data during knee extension, especially the stress during knee hyperextension. The measurement range is 0~50N, with a resolution of 0.1N. Each sensor is non-invasively attached to the human body surface using medical low-sensitivity adhesive or elastic sleeve. The posture sensor and discharge device at the knee joint sleeve can be integrated into a single design, reducing the number of components and improving wearing comfort. The elastic sleeve is available in S / M / L / XL sizes according to height and weight to ensure a close fit without compressing blood vessels.

[0007] The auxiliary stimulation module includes a medical-grade neuromuscular electrical stimulation (EMS) discharge device, which is placed only on the affected side of the user. Its placement follows the electrode placement principles for EMS, coinciding with the area collected by the surface electromyography (EMG) sensor (e.g., the corresponding areas of the quadriceps femoris and peroneus longus muscles). The core function of the discharge device is guided stimulation, not strong stimulation. Stimulation parameters are adjustable: intensity range 0.1~5mA (0.1mA steps), frequency range 1~10Hz, pulse width 200~500μs. It features overcurrent protection and skin impedance detection to prevent abnormal stimulation from damaging the skin. An emergency pause button is located on the side of the component, allowing the user to interrupt stimulation at any time.

[0008] The wireless communication module uses Bluetooth 5.0 or LoRa wireless networking technology to achieve bidirectional data transmission between the wearable sensing module, the auxiliary stimulation module, and the cloud data processing module. The communication distance is ≥10m, the latency is ≤50ms, and it supports simultaneous connection of up to 12 components, ensuring the timing consistency of data from each sensor. It also features signal strength monitoring, automatically reminding the user to adjust component positions or eliminate interference (such as metal obstructions or strong electromagnetic environments) when the signal strength drops below -70dBm.

[0009] The cloud-based data processing module employs an encrypted storage architecture (compliant with HIPAA medical data privacy standards) to store raw data collected by the wearable sensing module, personalized assessment model parameters, rehabilitation assessment results, and historical training data. It incorporates a built-in human motion standard model, using algorithms to compare real-time user data with standard parameter ranges to quantitatively analyze indicators such as the rationality of movement posture, muscle force symmetry, and the incidence of abnormal movements. It supports real-time or scheduled data synchronization (every 5 minutes) and features an abnormal data alert function, allowing notifications to be pushed to medical staff via the human-computer interaction module.

[0010] The human-computer interaction module includes a mobile app supporting iOS / Android systems and a Windows / web UI, with the following core functions: Device configuration: Supports wireless pairing of components, sampling rate adjustment, evaluation threshold setting, and stimulation parameter configuration; Data display: Real-time display of parameters such as joint angle, muscle force value, and force data, and display of historical data trends through line charts / bar charts; Motion Recreation: The built-in 3D human motion performance module maps the parameters collected by the sensors to the virtual character, and reproduces the user's movement process from multiple angles through animation; Report generation: Automatically generate rehabilitation progress reports on a weekly / monthly basis, including core assessment indicators, improvement of abnormal movements, and achievement rate of rehabilitation goals; Data sharing: Supports user-authorized access to data for medical staff and family members. Medical staff can remotely view real-time training data and adjust the plan via the web interface; Data Export: Supports exporting raw data and evaluation reports to CSV / Excel format, facilitating medical record archiving or scientific research analysis.

[0011] A rehabilitation treatment assessment method, the specific steps of which are as follows: Step S1: Equipment Pre-processing and Personalized Configuration Skin preparation: Users should clean the skin area in contact with the component with warm water and mild soap, and dry it to ensure that the skin is dry, undamaged, and free of hair (trim hair if necessary) to avoid affecting the sensor's fit and adhesive adhesion. Component Fixation and Inspection: Medical personnel or users fix the wearable sensing module and auxiliary stimulation module in the preset positions according to the Chinese labels and positioning diagram of the components. After completion, three verifications are performed: the user performs squatting and walking movements to check the stability of the components (no displacement or detachment); the user is asked to confirm that there is no stinging or pressure sensation (no lifting of the adhesive edges); the data of each sensor is viewed through the human-computer interaction module to confirm that there is no disconnection or abnormal value. Wireless pairing and calibration: Enable the wireless function of all components, initiate a search through the human-computer interaction module and complete automatic network pairing; Calibrate the sensors: The posture sensor completes horizontal calibration and joint zero-position calibration in the user's standing neutral position, the tension sensor is calibrated by zero load calibration and standard weight (5N / 10N / 20N) loading calibration (error ≤ ±2%), and the surface electromyography sensor is calibrated to the reference value by the user's maximum voluntary contraction (MVC) action; Personalized parameter settings: Medical staff set core parameters according to the user's rehabilitation scenario (such as hemiplegia after stroke, post-knee surgery) and physical condition: preset standard range of joint range of motion (such as knee flexion and extension 0°~135°), abnormal movement trigger threshold (such as knee hyperextension>15°), and initial stimulation parameters (intensity 0.5~1mA, frequency 3~5Hz).

[0012] Step S2: Adaptive Model Training (Personalized Model Building) New users need to complete a 3-day adaptation training under the full guidance of professional medical personnel. The core purpose is to allow the cloud data processing module to build a personalized assessment model for the user, while also allowing the user to adapt to the wearing comfort and stimulation mode of the device. After training, the cloud data processing module automatically generates a "Personalized Exercise Model Report for Users," which includes the user's baseline exercise parameters, abnormal movement types (such as knee hyperextension and foot drop), and stimulation parameter optimization plans, serving as the basis for subsequent rehabilitation assessment and treatment.

[0013] Step S3: Real-time Data Acquisition. Under the guidance of medical personnel, users conduct routine rehabilitation training (such as gait training, balance training, and joint range of motion training) or home-based self-training. The wearable sensing module simultaneously collects multi-dimensional data at a preset sampling rate: the posture sensor collects the relative position of joints, angular velocity, and inertial acceleration; the surface electromyography sensor collects the intensity of muscle electrical signals; and the force sensor collects the force data in the popliteal fossa. During the acquisition process, the wireless communication module uploads the raw data to the cloud data processing module in real time, and the human-computer interaction module displays key parameters simultaneously, allowing users to understand their own movement status in real time.

[0014] Step S4: Cloud-based quantitative assessment. The cloud data processing module completes the assessment according to the following process. Data preprocessing: The raw data is filtered to remove motion artifacts and environmental interference; Feature extraction: Extract core features such as joint range of motion, peak muscle force, peak force, and frequency of abnormal movements; Comparative analysis: The extracted features are compared with the standard parameter range and historical data in the personalized evaluation model; Indicator quantification: Generate quantitative assessment results, including: joint range of motion compliance rate (actual range of motion / target range of motion × 100%), muscle force symmetry coefficient (force value on the affected side / force value on the healthy side), frequency of abnormal movements (number of movements exceeding the threshold per unit time), and gait stability parameters (step length variation coefficient, gait frequency stability). Report generation: The assessment results are pushed to the human-computer interaction module in real time, and a rehabilitation progress report is generated by summarizing on a weekly / monthly basis.

[0015] Step S5: Multi-dimensional feedback and adjustment Visual feedback: Users, medical staff and their families can view the 3D animation motion reproduction and assessment results through the human-computer interaction module, and intuitively understand the shortcomings in the motion process (such as knee hyperextension and foot drop in stroke patients). Stimulation-guided feedback: When the cloud data processing module detects that the user's movement exceeds the abnormal threshold (such as knee hyperextension > 15°, foot eversion > 10°), it automatically sends a trigger command to the auxiliary stimulation module. The discharge device provides mild electrical stimulation according to preset parameters to remind the user to adjust the muscle exertion mode. As the user's rehabilitation progresses (the frequency of abnormal movements decreases by more than 50%), the system automatically reduces the stimulation frequency (such as from "triggering every abnormality" to "triggering once every 3 abnormalities") or increases the trigger threshold, gradually reducing device intervention and cultivating the user's autonomous motor control ability. Solution optimization: Based on historical data and assessment reports stored in the cloud, medical staff dynamically adjust the rehabilitation training plan, assessment thresholds, and stimulation parameters during monthly follow-up visits; if the user is not comfortable with the stimulation mode, the stimulation type (continuous pulse / intermittent pulse) can be switched.

[0016] The technical advantages of this invention are as follows: Through the collaborative work of wearable sensing modules, wireless communication, cloud data processing, human-computer interaction and auxiliary stimulation modules, and standardized rehabilitation treatment assessment process, a complete closed loop of "sensing-analysis-assessment-feedback" in rehabilitation treatment is realized. This not only makes up for the lack of quantitative indicators in cognitive motor therapy (CTE), but also solves the problem that existing sensors can only detect but cannot deeply assist in treatment. It collects parameters such as joint movement, muscle activity, and force from multiple dimensions, and combines cloud-based human movement models and personalized assessment models for precise quantitative analysis to generate objective and traceable rehabilitation data. This effectively reduces the subjective component of traditional rehabilitation assessment. At the same time, it uses 3D animation to reproduce movement and guided electrical stimulation to provide users with intuitive feedback and targeted adjustment guidance, and provides data support for medical staff to optimize rehabilitation plans. This system and method support home training and remote guidance, significantly reducing the manpower and time costs of conventional rehabilitation treatment. Furthermore, it can be adapted to various rehabilitation scenarios such as stroke hemiplegia and post-operative sports injury through parameter adjustment, ultimately achieving standardization, personalization, and efficiency in rehabilitation treatment, and greatly improving the pertinence, effectiveness, and scalability of rehabilitation treatment. Attached Figure Description

[0017] Figure 1 The diagram illustrates the specific steps of the method of this invention. Figure 2 This is a block diagram of the system of the present invention; Figure 3 This is a distribution diagram of the attitude sensors in this invention; Figure 4 This is a distribution diagram of the tension sensors in this invention; Figure 5 This diagram shows the symmetrical distribution of all sensors in this invention. Detailed Implementation

[0018] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0019] I. System Component Preparation and Selection Wearable sensing module selection: The posture sensor is the Wittek WT901C nine-axis module (sampling rate 100Hz, angle measurement range -180°~180°), the surface electromyography sensor is the DelsysTrigno Wireless dry electrode module (sampling rate 200Hz, input impedance ≥10MΩ), and the tension sensor is the TekscanFlexiforce thin film sensor (measurement range 0~50N, thickness 2mm). Assistive stimulation module selection: Select a medical-grade neuromuscular electrical stimulation module (stimulation intensity 0.1~5mA, frequency 1~10Hz), equipped with low-sensitivity electrode pads; Wireless communication module: adopts Bluetooth 5.0 technology and is equipped with a signal enhancement antenna; Human-computer interaction module: Install a custom-developed mobile APP and web UI backend, and import human motion standard model parameters in advance (such as knee flexion and extension 0°~135°, and popliteal fossa tension ≤30N).

[0020] II. Step S1: Equipment Pre-processing and Personalized Configuration Skin preparation: The patient cleans the skin around the left hip, knee, ankle, quadriceps, and popliteal fossa, and trims any excessive hair after drying to ensure the skin is dry and free of oil. Component fixed: Posture sensor: symmetrically fixed to the midpoint of the lateral side of the hip (near the greater trochanter of the femur), knee (above the head of the fibula), and ankle (above the tip of the lateral malleolus) joints on both sides using 3M1527C low-sensitivity adhesive. The adhesive force on the left side (affected side) is slightly greater than that on the healthy side. Tension sensor: cut into a 10cm×3cm strip, attach it to the midpoint behind the popliteal fossa, and fix both ends with elastic sleeves; Surface electromyography (EMG) sensors: attached to the quadriceps femoris (upper 1 / 3 of the anterior thigh), gastrocnemius (upper 1 / 3 of the posterior calf), peroneus longus (lower 1 / 3 of the lateral calf), and biceps femoris (upper 1 / 3 of the posterior thigh), with one sensor for each muscle group. The electrode pads are parallel to the direction of the muscle fibers. Discharge device: fixed in the corresponding area of ​​the left quadriceps femoris and peroneus longus muscles, overlapping with the surface electromyography sensor; Wearing test: The patient squats and walks slowly, and the components are observed to be in no position; the patient is asked if there is any pressure on the skin and if the adhesive edges are lifted; the data of each sensor is viewed through the APP to confirm that the joint angle, electromyography signal and tension value are collected normally; Wireless pairing and calibration: Open the APP to initiate pairing. All components will automatically form a network successfully. The APP will display the device number, battery level, and signal strength (≥-65dBm) of each component. Posture sensor calibration: The patient stands in a neutral position and follows the APP prompts to complete the horizontal calibration and joint zero-position calibration; Force sensor calibration: With the patient's knee joint extended, after completing zero-load calibration, load and calibrate sequentially with 5N, 10N, and 20N standard weights to ensure that the measurement error is ≤±2%; Surface electromyography sensor calibration: Under the guidance of medical staff, the patient performs the maximum voluntary contraction of the left quadriceps muscle 3 times (5 seconds each time), and the APP automatically records the peak value of the electromyography signal as a benchmark. Personalized parameter settings: Data acquisition parameters: posture sensor 100Hz, surface electromyography sensor 200Hz, tension sensor 50Hz; Assessment thresholds: knee flexion and extension range 0°~135°, knee hyperextension threshold >15°, popliteal fossa traction threshold >30N, foot eversion threshold >10°; Stimulation parameters: intensity 0.8mA, frequency 4Hz, pulse width 300μs, trigger conditions bound to knee hyperextension >15° and foot eversion >10°.

[0021] III. Step S2: Adaptive Model Training (3 days) Day 1 (Static Movement Training): Training content: Standing neutral position (10 seconds x 5 sets), seated knee flexion / extension (0°~90°, 10 seconds x 5 sets), ankle dorsiflexion / plantar flexion (10 seconds x 5 sets); Medical staff intervention: real-time monitoring of APP data, adjustment of the position of the left ankle joint posture sensor (initial misalignment caused abnormal data); observation of the patient's skin for any allergic reactions; Model output: Generates a static motion baseline model of the patient in the cloud, recording the baseline values ​​of bilateral joint range of motion and muscle force baseline values.

[0022] Day 2 (Dynamic Basic Movement Training): Training content: Slow walking (10 meters x 3 sets), going up and down stairs (5 steps x 3 sets), slow turns (5 times each to the left and right); Medical staff intervention: When the patient exhibited left knee hyperextension (approximately 18°) while walking, the APP triggered a stimulus, and the patient reported that the stimulus intensity was moderate; the stimulus trigger delay was fine-tuned (from 0.5 seconds to 0.3 seconds) to ensure timely reminders; Model optimization: Personalized models are updated in the cloud, and the dynamic action recognition algorithm is optimized, improving the abnormal action recognition rate to 96%.

[0023] Day 3 (Simulated daily activity training): Training content: dressing (with left upper limb assistance), retrieving an object (a water cup 50cm in front), short-distance walking (50 meters), sitting up; Medical staff intervention: Teach patients' families how to view 3D motion reproduction and basic data through the APP; verify the model's accuracy in recognizing abnormal movements in daily activities (100% recognition rate for knee hyperextension and foot eversion). Model output: Generates a "Personalized Movement Model Report for Patients", which identifies left knee hyperextension (average 16.2°) and foot eversion (average 12.5°) as the core abnormal movements, with a muscle force symmetry coefficient of 0.6 (left quadriceps force is 60% of that of the healthy side).

[0024] IV. Step S3: Real-time Data Acquisition The patient performs gait rehabilitation training at home (30 minutes daily, divided into 3 sets). During the training: The posture sensor collects the relative position and angular velocity of the bilateral hip, knee and ankle joints in real time, and captures the knee hyperextension (up to 17.3°) in the left support phase and the foot eversion (up to 13.1°) in the swing phase. Surface electromyography (EMG) sensors were used to collect EMG signals from the left quadriceps femoris (peak value 40% lower than the healthy side) and data on uneven force exertion in the gastrocnemius muscle. The tension sensor recorded a maximum tension of 32.6 N in the left popliteal fossa when the knee was extended (exceeding the threshold of 30 N). All data is uploaded to the cloud in real time via Bluetooth 5.0, and the APP homepage displays joint angle curves, muscle force bar graphs, and tension values ​​in real time.

[0025] V. Step S4: Cloud-based quantitative assessment The cloud data processing module analyzes data according to the following process: Data preprocessing: filtering out two abnormal data points caused by slight sensor displacement during the walking process; Feature extraction: Frequency of left knee hyperextension (12 times / 30 minutes), frequency of foot eversion (8 times / 30 minutes), quadriceps muscle force symmetry coefficient (0.62), and number of times popliteal fossa pull exceeded the standard (5 times). Comparative analysis: Compared with the baseline values ​​in the personalized model, the frequency of knee hyperextension decreased by 18% and the frequency of foot eversion decreased by 15% compared with the third day of training, while the force symmetry coefficient increased by 3.3%. Assessment results: A real-time assessment report was generated, marked as "significant improvement in knee hyperextension, moderate improvement in foot eversion, and room for improvement in muscle symmetry". The rehabilitation goal achievement rate was 35% (weekly goal: 50% reduction in the frequency of abnormal movements). Warning notification: The APP sends a warning notification to medical staff because the popliteal fossa tension exceeds the standard three times in a row.

[0026] VI. Step S5: Multi-dimensional Feedback and Adjustment Visual feedback: Patients and their families can view 3D animations through the app to clearly observe when left knee hyperextension and foot eversion occur; they can also view weekly data trend charts to understand the improvement of abnormal movements; Stimulation-guided feedback: During training, when knee hyperextension >15° or foot eversion >10° is detected, the discharge device triggers a 0.8mA, 4Hz guided stimulus. After the patient feels a slight numbness, they can adjust their gait to reduce the knee hyperextension angle to within 10° and the foot eversion angle to within 8°. In the later stages of training (after the 20th minute), the frequency of abnormal movements decreases, and the system automatically adjusts the stimulation frequency to "trigger once every 2 abnormal movements". Solution optimization: After receiving the alert, medical staff can view the patient's real-time data and assessment report through the web UI and remotely adjust the parameters: lower the popliteal fossa tension threshold to 28N, fine-tune the stimulation intensity to 0.9mA, and add a "left gastrocnemius muscle stimulation node"; recommend that patients increase single-leg standing balance training (10 minutes per day) and send training guidance videos through the APP.

[0027] VII. Equipment Maintenance and Precautions Routine maintenance: After daily training, wipe the sensor probe and discharge device electrode with medical alcohol wipes (avoid alcohol seeping into the equipment); hand wash the elastic sleeve with warm water once a week (do not use bleach) and air dry naturally; replace the adhesive every 2 days. If slight redness or swelling occurs on the skin, replace it with hypoallergenic silicone adhesive. Power management: Check the component's power level before daily training. Wirelessly charge when the remaining power is less than 20% (charging time is 2-3 hours); charge once a month if not used for a long period of time. Safety Contraindications: This device is contraindicated for patients with broken or infected skin, allergies to adhesives / metals, or implanted pacemakers. If allergic reactions such as itching or redness occur during training, immediately remove the device, clean the skin with warm water, and apply an anti-allergy ointment if necessary. Troubleshooting: If there is no data transmission, check the component battery level in the following order: → restart the device and re-pair → eliminate wireless interference → contact after-sales service for testing. Figure 3 In the diagram: the orange module is a posture sensor, the green module is a surface electromyography (EMG) acquisition device, the blue module is a discharge device, and the black module is an elastic sleeve or underwear. The green or orange module can be attached to the body surface using adhesive. The sensor and discharge module at the knee joint sleeve can be integrated. Figure 4 In the middle: the yellow and white modules are tension sensors, used to sense the movement of the knee joint during walking, mainly the stress during knee hyperextension; Figure 5 In the middle: all sensors are symmetrically distributed, but the discharge device is only placed on the affected side (taking a patient with left hemiplegia as an example).

[0028] This embodiment, through standardized assessment methods and an integrated system, achieves quantitative assessment and personalized intervention for the rehabilitation treatment of stroke patients. Patients can intuitively understand their own motor deficiencies through 3D animation and stimulus feedback. Medical staff can dynamically optimize the treatment plan based on objective data, significantly improving the pertinence and efficiency of rehabilitation treatment. After four weeks of continuous use, the frequency of left knee hyperextension decreased by 72%, the frequency of foot eversion decreased by 68%, the muscle force symmetry coefficient increased to 0.85, and the rehabilitation goal achievement rate was 92%, verifying the effectiveness and practicality of the invention.

[0029] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A rehabilitation treatment assessment system based on wearable devices, characterized in that, include: The wearable sensing module includes posture sensors symmetrically distributed on the main joints of the hip, knee, and ankle; surface electromyography sensors attached to the main muscle groups such as the quadriceps femoris; and a tension sensor located behind the popliteal fossa, for synchronously collecting joint motion parameters, muscle activity parameters, and force parameters. The auxiliary stimulation module includes a discharge device that is placed only on the affected side of the user to provide guided electrical stimulation without generating strong stimulation; The wireless communication module adopts a wireless networking method to realize data transmission between the wearable sensing module, the auxiliary stimulation module and the cloud data processing module; The cloud-based data processing module is used to store the collected data, perform quantitative analysis on the data based on human movement models and personalized assessment models, and generate rehabilitation assessment results. The human-computer interaction module, including supporting software and web UI, supports device configuration, parameter adjustment, 3D motion animation reproduction, and evaluation report display for users, medical staff, and their families to view.

2. The rehabilitation treatment assessment system based on wearable devices according to claim 1, characterized in that, The sensors of the wearable sensing module are non-invasively attached to the human body surface using medical low-sensitivity adhesive or elastic sleeves. The posture sensor and discharge device at the knee joint sleeve can be integrated into a single design. The wearable sensing module supports commercially available mature modules or custom-developed integrated modules, and the sampling rate of each sensor is adjustable.

3. The rehabilitation treatment assessment system based on wearable devices according to claim 1, characterized in that, The wireless communication module uses Bluetooth 5.0 or LoRa technology, with a communication latency of ≤50ms, and supports simultaneous transmission across multiple devices; the cloud data processing module uses encrypted storage, complies with medical data privacy standards, and supports abnormal data alert push.

4. The rehabilitation treatment assessment system based on wearable devices according to claim 1, characterized in that, The placement of the discharge device in the auxiliary stimulation module is based on the principles of neuromuscular electrical stimulation electrode placement and is adjusted according to the user's actual motor performance. The electrical stimulation parameters include an intensity of 0.1~5mA and a frequency of 1~10Hz, which can be set by medical personnel or artificial intelligence intervention through the human-computer interaction module.

5. A rehabilitation treatment assessment method based on wearable devices, characterized in that, Includes the following steps: S1: Equipment pretreatment and personalized configuration: Clean the skin of the human body contact area, fix the wearable sensing module in the preset position, complete the wireless networking pairing through the human-computer interaction module, and medical staff set data collection parameters, assessment thresholds and initial stimulation parameters according to the user's rehabilitation needs. S2: Adaptive Model Training: New users undergo 3 days of adaptive training under the guidance of professional medical personnel. The wearable sensing module continuously collects static and dynamic motion data, and the cloud data processing module builds a personalized evaluation model for users based on the collected data, optimizing the evaluation threshold and stimulus triggering conditions. S3: Real-time data acquisition: When the user is performing rehabilitation training, the wearable sensing module synchronously collects the relative position of the joint, angular velocity, inertial acceleration, muscle electrical signals and knee extension force data, and uploads them to the cloud data processing module in real time through the wireless communication module. S4: Cloud-based quantitative assessment: The cloud-based data processing module compares the collected data with the standard parameter range in the personalized assessment model, quantitatively analyzes the rationality of the movement posture, the symmetry of muscle exertion, and the incidence of abnormal movements, and generates real-time rehabilitation assessment results. S5: Multi-dimensional feedback and adjustment: The human-computer interaction module reproduces the user's movement process through 3D virtual character animation and displays the evaluation results simultaneously; when abnormal movement is detected, the auxiliary stimulation module triggers guided electrical stimulation according to preset parameters to remind the user to adjust their movements; Medical staff dynamically optimize rehabilitation training programs based on historical data and assessment reports stored in the cloud.

6. The rehabilitation treatment assessment method based on wearable devices according to claim 5, characterized in that, In step S1, the data acquisition parameters include: posture sensor sampling rate ≥100Hz, surface electromyography sensor sampling rate ≥200Hz, and tension sensor sampling rate ≥50Hz; the evaluation threshold is based on a preset human motion model and can be manually adjusted according to the user's physical condition.

7. The rehabilitation treatment assessment method based on wearable devices according to claim 5, characterized in that, The adaptive training in step S2 includes: Day 1: static motion training, collecting basic motion parameters to establish a baseline model; Day 2: dynamic basic motion training, optimizing the model's recognition accuracy for complex movements; Day 3: simulated daily activities training, improving the personalized evaluation model, and generating a personalized motion model report for the user after training.

8. The rehabilitation treatment assessment method based on wearable devices according to claim 5, characterized in that, The quantitative assessment indicators in step S4 include: joint range of motion achievement rate, frequency of abnormal movements, muscle force symmetry coefficient, and gait stability parameters. The assessment results are presented in numerical and graphical form, and support the generation of rehabilitation progress trend analysis on a weekly / monthly basis.

9. The rehabilitation treatment assessment method based on wearable devices according to claim 5, characterized in that, In step S5, the electrical stimulation triggering conditions of the auxiliary stimulation module are bound to the abnormal movement threshold. As the user's rehabilitation progresses, the stimulation frequency is automatically reduced or the triggering threshold is increased. The human-computer interaction module supports data export to CSV / Excel format for medical record archiving or scientific research analysis.

Images