A stroke rehabilitation evaluation system based on multi-information fusion
By combining a central processing belt and AR smart glasses with a multi-information fusion stroke rehabilitation assessment system, the problems of subjective bias and data security in traditional assessments have been solved, achieving efficient and safe rehabilitation assessment.
Patent Information
- Application Number
- CN202611487544.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-09-23
- Publication Date
- 2026-11-10
AI Technical Summary
Traditional stroke rehabilitation assessment relies on manual observation, which is subject to subjective bias and data transmission delays, and cannot achieve continuous dynamic data collection. Furthermore, intelligent assessment devices pose data security risks.
It employs a central processing belt, a sensor computing module, and AR smart glasses, integrating elbow and lower limb motion detection units and electromyography detection units for local data processing and analysis. It communicates with the belt through a multi-mode transmission module and uses encryption technology to ensure data security.
It enables precise and personalized rehabilitation assessments, reduces data transmission volume, improves the security and accuracy of assessments, and minimizes the impact on user experience.
Smart Images

Figure CN123058723A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of rehabilitation assessment technology, and in particular to a stroke rehabilitation assessment system based on multi-information fusion. Background Technology
[0002] Stroke is a common acute cerebrovascular disease. Even with timely treatment, approximately 70% of survivors will experience varying degrees of motor dysfunction, cognitive impairment, and psychological and emotional abnormalities. This not only severely reduces the patient's quality of life but also places a heavy medical burden on families and society. Rehabilitation assessment is a core component of the entire stroke rehabilitation process. Accurate and personalized assessment results are the foundation for developing targeted rehabilitation plans, tracking progress, and adjusting treatment strategies. The comprehensiveness and accuracy of the assessment results directly determine the final outcome of rehabilitation treatment.
[0003] Clinical scale-based assessment is the mainstream assessment method for traditional stroke rehabilitation. It relies on the manual observation of rehabilitation physicians. Not only are the assessment results easily affected by the physician's subjective experience and the fluctuation of the patient's condition on the day of the test, but it can also only capture the patient's functional state at a static point in time. It cannot achieve continuous dynamic collection of various physiological and motor parameters during the patient's daily rehabilitation training, and it is difficult to reflect the patient's true rehabilitation level.
[0004] To address the problems of traditional stroke rehabilitation assessment, a sensor-based intelligent rehabilitation assessment device is proposed. Existing intelligent rehabilitation assessment devices require transmitting sensor data to a cloud platform for analysis and assessment on the server side. This results in data transmission delays and data leakage, significantly reducing user experience and affecting assessment effectiveness. Summary of the Invention
[0005] To address the security risks associated with long-distance transmission of sensor data in existing technologies, this invention proposes a stroke rehabilitation assessment system based on multi-information fusion, comprising a central processing belt, a sensing and computing module, and AR smart glasses. The central processing belt is communicatively connected to the sensing and computing module and the AR smart glasses, respectively.
[0006] The sensing and processing module includes an elbow motion detection unit, a lower limb motion detection unit, and an electromyography (EMG) detection unit. The elbow motion detection unit includes motion angle detection components worn on the patient's left and right wrists, respectively. The lower limb motion detection unit includes motion angle detection components worn on the patient's left and right ankles, respectively. The EMG detection unit includes an EMG signal processing component and an EMG signal acquisition component that is attached to the surface of the biceps brachii, triceps brachii, and quadriceps femoris muscles via a strap.
[0007] The AR smart glasses include an interaction module, a shooting module, a heart rate detection module, a display module, and a control module;
[0008] The computing belt includes a belt body, a main processor located in the center of the belt body, and a main storage module and a communication module located on both sides of the main processor.
[0009] Furthermore, the motion angle detection unit includes an integrated detection module, a storage module, a first processing chip, and a first multi-mode transmission module. The integrated detection module is a low-power 6-axis inertial measurement unit BMI160.
[0010] Furthermore, the electromyography signal acquisition component includes an electromyography detection module, a second control chip, and a second transmission module; the electromyography signal processing component includes a third processing chip and a third multi-mode transmission module; and the third multi-mode transmission module includes an information input chip and an information output chip.
[0011] Furthermore, the shooting module includes an environmental perception camera and an eye-tracking camera, the heart rate detection module uses a PPG sensor installed in the nose pad of the glasses, and the control module includes a central controller and a fourth multi-mode transmission module.
[0012] Furthermore, based on a comprehensive rehabilitation assessment score, the stroke rehabilitation assessment system performs automated rehabilitation assessments using sensor data.
[0013] Furthermore, the comprehensive rehabilitation assessment scoring items include elbow flexion, wrist circular motion, knee flexion in the supine position, knee extension in the supine position, knee flexion in the sitting position, muscle fatigue level, and emotional regulation ability. Each item is scored as follows: 0 points, 1 point, and 2 points. 0 points indicates that the error between the test data and the baseline value is greater than 20%, 1 point indicates that the error between the test data and the baseline value is less than or equal to 20% and greater than 5%, and 2 points indicates that the error between the test data and the baseline value is less than 5%.
[0014] Furthermore, the stroke rehabilitation assessment system uses an elbow motion detection unit to assess elbow flexion, AR smart glasses to assess wrist circular motion, spherical grasp and emotion regulation ability, a lower limb motion detection unit to assess knee flexion and knee extension, and an electromyography detection unit to assess muscle fatigue.
[0015] Furthermore, a single assessment process specifically includes initial setup, followed by a health assessment in the order of elbow flexion, wrist circular motion, knee flexion in a supine position, knee extension in a supine position, and knee flexion in a seated position. The elbow motion detection unit, AR smart glasses, and lower limb motion detection unit transmit the project scores to the main processor. During the health assessment, the electromyography (EMG) detection unit calculates the score for muscle fatigue, and the AR smart glasses calculate the score for emotional regulation ability. The EMG detection unit and AR smart glasses transmit the calculated scores to the main processor, which then aggregates them to obtain a comprehensive rehabilitation assessment score.
[0016] Furthermore, the initialization settings specifically include setting the movement patterns and durations for elbow flexion, wrist circular motion, knee flexion in a supine position, knee extension in a supine position, and knee flexion in a seated position. The movement patterns include unilateral movement and bilateral synchronous movement.
[0017] Furthermore, at 11 PM every day, the motion angle detection component encrypts the stored angle information through the first processing chip, connects two motion angle detection components to the central processing belt via 5G Wi-Fi, and connects two motion angle detection components to the central processing belt via Bluetooth. The motion angle detection components transmit the encrypted angle information to the main storage module of the central processing belt. After the transmission is completed, the motion angle detection components delete the encrypted angle information. The main processor organizes the transmitted data to obtain the data of each evaluation process of the day, and re-encrypts the data of each evaluation process.
[0018] Compared with the prior art, the present invention has the following beneficial effects:
[0019] During the evaluation process, the sensing and computing module and the AR smart glasses complete all data processing and analysis. The original sensing data cannot be directly reconstructed from the processing results. The central processing belt only summarizes the processing results, which reduces the amount of data transmission while ensuring data security. Furthermore, after the evaluation process is completed, the motion angle detection data obtained by the sensing and computing module is stored in the central processing belt with two layers of encryption to further ensure data security. Attached Figure Description
[0020] Figure 1 This is a schematic diagram of the stroke rehabilitation assessment system of the present invention. Detailed Implementation
[0021] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.
[0022] like Figure 1 As shown, the stroke rehabilitation assessment system based on multi-information fusion proposed in this invention includes a central processing belt, a sensing and computing module, and AR smart glasses. The central processing belt is connected to the sensing and computing module and the AR smart glasses for communication.
[0023] The sensing and processing module includes an elbow motion detection unit, a lower limb motion detection unit, and an electromyography (EMG) detection unit. The elbow motion detection unit includes motion angle detection components worn on the patient's left and right wrists, respectively. The lower limb motion detection unit includes motion angle detection components worn on the patient's left and right ankles, respectively. Each motion angle detection component has a ring-shaped structure, including an integrated detection module, a storage module, a first processing chip, and a first multi-mode transmission module. All motion angle detection components use the same processor for their first processing chip. The integrated detection module uses the Bosch BMI160 low-power 6-axis inertial measurement unit, which includes a three-axis accelerometer and a three-axis gyroscope. The first processing chip is the Texas Instruments AM62Ax processor. The first multi-mode transmission module uses the NXP 88W8897, which supports 2.4 / 5GHz dual-band Wi-Fi and Bluetooth 4.2. The EMG detection unit includes an EMG signal processing component and six EMG signal acquisition components. The electromyography (EMG) signal acquisition component includes an EMG detection module, a second control chip, and a second transmission module. The EMG detection module uses a Shimmer3 EMG sensor, the second control chip is a Texas Instruments MSP430G2201IPW14R microcontroller, and the second transmission module uses an Espressif ESP32-D0WD-V3. The EMG signal processing component includes a third processing chip and a third multi-mode transmission module. The third processing chip is also a Texas Instruments AM62Ax processor. The third multi-mode transmission module includes an input chip and an output chip. The input chip communicates with each EMG detection unit and uses an ANS-BT105M Bluetooth module from Anshuo Technology, supporting up to 3 master and 3 slave devices, allowing for stable simultaneous connection to 6 devices. The output chip communicates with the central processing belt and uses an NXP 88W8897. The electromyography (EMG) signal acquisition components are respectively set on the surface of the left biceps brachii, right biceps brachii, left triceps brachii, right triceps brachii and left quadriceps femoris via straps. The EMG signal acquisition components and EMG signal processing components are respectively set on the surface of the right quadriceps femoris via straps.
[0024] The AR smart glasses include an interaction module, a camera module, a heart rate detection module, a display module, and a control module. The control module communicates with the interaction module, camera module, heart rate detection module, and display module. The interaction module includes a speaker and a microphone array. The speaker uses AAC Technologies' AR-specific super-linear speaker. The microphone array includes six microphones, with three integrated into each temple. The microphones are all integrated into the inside of the temple, close to the user's face. In each temple, two microphones are close to the frame to collect sound from the front, and one microphone is close to the end of the temple to collect the wearer's own voice. The microphones are MEMS microphones MSM381ACB026 from Suzhou Minxin Microelectronics Technology Co., Ltd. The camera module includes an environmental awareness camera and an eye-tracking camera. The environmental awareness camera is integrated into the center of the frame above the nose, providing a view similar to that of the human eye, resulting in images that closely resemble what the human eye actually sees. The environmental awareness camera uses a Sony IMX686 sensor. The eye-tracking cameras are embedded in the frame, with two cameras for each eye to accurately identify eye movements and pupil changes. The eye-tracking cameras use an OmniVision OV7251 sensor. The heart rate monitoring module uses a PPG sensor located in the nose pad of the glasses. The PPG sensor emits light to illuminate the skin, detecting changes in light absorption / scattering caused by blood flow, capturing signals generated by blood volume pulsation, and calculating the heart rate value for continuous monitoring. The display module uses a Sony 0.68-inch Micro OLED display. The control module includes a central controller and a fourth multi-mode transmission module. The central controller uses a Snapdragon XR2+ Gen 2 processor to control the operation of the AR smart glasses, while the fourth multi-mode transmission module uses an NXP 88W8897 processor.
[0025] The central processing belt consists of a belt body made of elastic, breathable, and skin-friendly material. A main processor is located in the center of the belt body, with a main storage module and a communication module on either side of the main processor. The main processor is the Jetson Orin Nano from NVIDIA, used to control the operation of the stroke rehabilitation assessment system. The communication module includes a Bluetooth unit and a Wi-Fi unit. The Bluetooth unit is the ANS-BT105M Bluetooth module from Anshuo Technology, and the Wi-Fi unit is the Espressif ESP32-C5. The Espressif ESP32-C5 supports 2.4GHz / 5GHz dual-band Wi-Fi and can stably support 5-8 devices simultaneously in 5GHz AP mode.
[0026] Referring to the Fugl-Meyer motor function score and considering actual sensor data, a comprehensive rehabilitation assessment score was constructed. The comprehensive rehabilitation assessment score includes items such as elbow flexion, wrist circular motion, knee flexion in the supine position, knee extension in the supine position, knee flexion in the sitting position, muscle fatigue level, and emotional regulation ability. Each item is scored as 0, 1, or 2 points. 0 points indicates an error greater than 20% between the measured data and the baseline value; 1 point indicates an error less than or equal to 20% and greater than 5% between the measured data and the baseline value; and 2 points indicate an error less than 5% between the measured data and the baseline value. A comprehensive rehabilitation assessment score less than 7 points is classified as Level 1, indicating very low rehabilitation progress requiring intensive training; a score greater than or equal to 7 points and less than 11 points is classified as Level 2, indicating significant progress requiring continued effort; a score greater than or equal to 11 points and less than 14 points is classified as Level 3, indicating basic recovery; and a score of 14 points indicates complete recovery.
[0027] Based on a comprehensive rehabilitation assessment score, the stroke rehabilitation assessment system utilizes an elbow motion detection unit to assess elbow flexion, AR smart glasses to assess wrist circular motion, spherical grasp, and emotion regulation ability, a lower limb motion detection unit to assess knee flexion and extension, and an electromyography (EMG) detection unit to assess muscle fatigue. A single assessment process includes initial setup, followed by a health assessment in the sequence of elbow flexion, wrist circular motion, knee flexion in a supine position, knee extension in a supine position, and knee flexion in a seated position. The elbow motion detection unit, AR smart glasses, and lower limb motion detection unit transmit the scores to the main processor. During the health assessment, the EMG detection unit calculates the muscle fatigue score, and the AR smart glasses calculate the emotion regulation ability score. Both the EMG detection unit and AR smart glasses transmit their calculated scores to the main processor, which then aggregates them to obtain the comprehensive rehabilitation assessment score.
[0028] The initialization settings specifically include setting the movement patterns and durations for elbow flexion, wrist circular motion, knee flexion in the supine position, knee extension in the supine position, and knee flexion in the seated position. The movement patterns include unilateral movement and bilateral synchronous movement.
[0029] For elbow flexion, the interactive module of AR smart glasses explains the key points of the movement to the patient. While explaining the key points, the connection method between the elbow motion detection unit and the central processing belt is set according to the movement mode. Specifically, if the movement mode is unilateral, the motion angle detection component worn on the wrist of the motion sensor communicates with the central processing belt via Bluetooth. If the movement mode is bilateral synchronous movement, the motion angle detection components worn on the patient's left and right wrists communicate with the central processing belt via Bluetooth. After the patient responds that they understand the key points of the movement, they are reminded to start the exercise, and the timer is started, recording the specific start time according to the international standard ISO. In the pure time format of the 8601 format, the motion angle detection component acquires accelerometer and gyroscope data. The accelerometer and gyroscope data are fused through complementary filtering to obtain and record the elbow flexion angle. The specific calculation process is existing technology and will not be elaborated here. When the timer reaches the set exercise time, the patient is reminded to stop exercising. The specific start time and each elbow flexion angle are stored as a data entry, with the specific start time serving as the data index. The exercise process is divided into various exercise cycles based on the elbow flexion angle, and the duration of each cycle is calculated. The average of the durations is taken as the final exercise duration. The final exercise duration is compared with the corresponding baseline value, and a score is obtained based on the error. The score is transmitted to the main processor. When the main processor receives one score, it directly uses that score as the elbow flexion item score. When it receives two scores, it uses the average of the two scores as the elbow flexion item score.
[0030] For wrist-circle exercises, the AR smart glasses' interactive module introduces the key points of the movements and the exercise mode to the patient. The AR smart glasses communicate with the central processing belt via Bluetooth. The patient bends their head so that the glasses are aligned with their wrist. After the patient responds that they understand the key points of the movements and the exercise mode, they are prompted to start exercising, and the timer begins. The AR smart glasses' environmental perception camera captures the wrist movement process and transmits the video to the central controller. The central controller calculates the movement trajectory of each wrist frame by frame using the inter-frame difference method. The specific calculation process is existing technology and will not be elaborated here. When the timer reaches the set exercise time, the patient is prompted to stop exercising. The exercise process is divided into various exercise cycles based on the movement trajectory of each wrist, and the exercise duration of each cycle is calculated. The average of the exercise durations is taken as the final exercise duration for each wrist. The final exercise duration is compared with the corresponding benchmark value, and a score is obtained for each wrist based on the error. If there is only one wrist, the score is used as the item score for the wrist-circle exercise and transmitted to the main processor. If there are two wrists, the average of the two scores is used as the item score for the wrist-circle exercise and transmitted to the main processor.
[0031] For knee flexion in the supine position, knee extension in the supine position, and knee flexion in the sitting position, the AR smart glasses' interactive module explains the key points of the movements to the patient. While explaining the key points, the connection method between the lower limb motion detection unit and the central processing belt is set according to the exercise mode. Specifically, if the exercise mode is unilateral, the motion angle detection component worn on the ankle of the motion detector communicates with the central processing belt via Bluetooth; if the exercise mode is bilateral synchronous exercise, the motion angle detection components worn on the patient's left and right ankles respectively communicate with the central processing belt via Bluetooth. After the patient responds that they understand the key points of the movements, they are reminded to start exercising, and the timer begins, recording the specific start time according to the international standard ISO. In the pure time format of the 8601 format, the motion angle detection component acquires accelerometer and gyroscope data. Complementary filtering is used to fuse the accelerometer and gyroscope data to obtain and record the knee joint angle. The specific calculation process is existing technology and will not be elaborated here. When the timer reaches the set exercise time, the patient is reminded to stop exercising. The specific start time and each knee joint angle are stored as a data entry, with the specific start time serving as the data index. The exercise process is divided into various exercise cycles based on the knee joint angles, and the duration of each cycle is calculated. The average of the durations is taken as the final exercise duration. The final exercise duration is compared with the corresponding baseline value, and a score is obtained based on the error. This score is transmitted to the main processor. When the main processor receives a score, it directly uses that score as the item score for knee flexion in the supine position, knee extension in the supine position, or knee flexion in the sitting position. When it receives two scores, the average of the two scores is used as the item score for knee flexion in the supine position, knee extension in the supine position, or knee flexion in the sitting position.
[0032] Throughout the exercise, during each exercise session, the muscles in a working state are identified. The electromyography (EMG) signal processing component communicates with the corresponding EMG signal acquisition component of the working muscles via Bluetooth. The EMG signal acquisition component transmits data to the EMG signal processing component. After all exercise sessions are completed, the EMG signal processing component calculates the integrated EMG values of the biceps brachii, triceps brachii, and quadriceps femoris. The specific calculation process is existing technology and will not be elaborated here. The average integrated EMG value of the three integrated EMG values is calculated and compared with the corresponding baseline value. Based on the error, a score for muscle fatigue is obtained. The EMG signal processing component communicates with the central processing belt via Bluetooth and transmits the calculated muscle fatigue score to the main processor.
[0033] Throughout the exercise, the AR smart glasses capture changes in eye features using an eye-tracking camera and transmit the data to the central controller. The central controller calculates the area of the pupil covered by the eyelids, and when the area is below a preset threshold, it is considered a blink. Data from the PPG sensor is also transmitted to the central controller, which separates the pulsation signal and performs spectral analysis on the pulsation signal using a Fast Fourier Transform to determine the fundamental frequency of the pulsation signal. The fundamental frequency is multiplied by 60 to obtain the heart rate per minute. After all the exercise events are completed, the number of blinks per minute is determined based on the total number of blinks and the total exercise time. The average heart rate during the entire exercise period is calculated, and the ratio of the number of blinks per minute to the average heart rate is compared with the corresponding baseline value. Based on the error, a score for emotion regulation ability is obtained. The AR smart glasses communicate with the central processing belt via Bluetooth and transmit the calculated emotion regulation ability score to the main processor.
[0034] At 11 PM every day, the motion angle detection component encrypts the stored angle information through the first processing chip. Two motion angle detection components communicate with the central processing belt via 5G Wi-Fi and Bluetooth. The motion angle detection components transmit the encrypted angle information to the main storage module of the central processing belt. After the transmission is completed, the motion angle detection components delete the encrypted angle information. The main processor organizes the transmitted data to obtain the data of each evaluation process of the day and re-encrypts the data of each evaluation process.
[0035] The specific process of encrypting the stored angle information is as follows: the specific start time is combined with the specific processor model of the first processing chip as the key base information. The first processing chip uses the SHA-256 algorithm to convert the key base information into a fixed 256-bit hash value as the key. Based on the key and the AES-256 algorithm, the angle data of each knee joint or each elbow flexion angle corresponding to the specific start time is encrypted.
[0036] The main processor organizes the transmitted data to obtain the data for each evaluation process of the day. The specific process is as follows: the angle data is sorted according to the specific start time of the angle data, and the angle data is grouped according to the required duration of one evaluation process. The result of subtracting the specific start time of the first data from the specific start time of the last data in each group of angle data is compared with the result of subtracting the set exercise duration of knee flexion in a seated position from the required duration of one evaluation process. The error is no more than 1 minute. Each group of angle data is the data for one evaluation process. The specific date and sequence number of the day are combined as the data index of the evaluation process. The sequence number represents the order of the evaluation process corresponding to each group of angle data in all evaluation processes of the day.
[0037] The specific process of re-encrypting the data in the evaluation process is as follows: the data index of the evaluation process and the specific processor model of the main processor are combined as the key base information. The main processor uses the SHA-256 algorithm to convert the key base information into a fixed 256-bit hash value. The first 64 bits and the last 64 bits of the hash value are combined as the key. The data in the evaluation process is encrypted based on the key and the AES-128 algorithm.
[0038] When an external terminal needs to obtain evaluation process data stored in the central processing unit, it needs to send a query request to the main processor. The main processor performs security verification on the query request. After successful verification, it determines the evaluation process to be queried based on the query request. It combines the data index of the evaluation process with the specific processor model of the main processor as the key information. The main processor uses the SHA-256 algorithm to convert the key information into a fixed 256-bit hash value. It combines the first 64 bits and the last 64 bits of the hash value as the key. Based on the key and the AES-128 algorithm, it decrypts the evaluation process data to obtain multiple angle data. It then sends a query request to the motion angle detection component to obtain the specific processor model of the first processing chip. The main processor combines the data index of each angle data with the specific processor model of the first processing chip as the key information. It uses the SHA-256 algorithm to convert the key information into a fixed 256-bit hash value as the key for each angle data. Based on the key and the AES-256 algorithm, it decrypts each angle data, achieving complete decryption of the evaluation process data. The fully decrypted evaluation process data is then fed back to the external terminal.
[0039] The embodiments described above are merely illustrative of specific implementations of the present invention, and while the descriptions are detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention.
Claims
1. A stroke rehabilitation assessment system based on multi-information fusion, characterized in that: The stroke rehabilitation assessment system includes a central processing belt, a sensing and computing module, and AR smart glasses. The central processing belt is communicatively connected to the sensing and computing module and the AR smart glasses, respectively. The sensing and processing module includes an elbow motion detection unit, a lower limb motion detection unit, and an electromyography (EMG) detection unit. The elbow motion detection unit includes motion angle detection components worn on the patient's left and right wrists, respectively. The lower limb motion detection unit includes motion angle detection components worn on the patient's left and right ankles, respectively. The EMG detection unit includes an EMG signal processing component and an EMG signal acquisition component that is attached to the surface of the biceps brachii, triceps brachii, and quadriceps femoris muscles via a strap. The AR smart glasses include an interaction module, a shooting module, a heart rate detection module, a display module, and a control module; The computing belt includes a belt body, a main processor located in the center of the belt body, and a main storage module and a communication module located on both sides of the main processor.
2. The stroke rehabilitation assessment system according to claim 1, characterized in that: The motion angle detection unit includes an integrated detection module, a storage module, a first processing chip, and a first multi-mode transmission module. The integrated detection module uses a low-power 6-axis inertial measurement unit BMI160.
3. The stroke rehabilitation assessment system according to claim 2, characterized in that: The electromyography (EMG) signal acquisition component includes an EMG detection module, a second control chip, and a second transmission module. The EMG signal processing component includes a third processing chip and a third multi-mode transmission module. The third multi-mode transmission module includes an information input chip and an information output chip.
4. The stroke rehabilitation assessment system according to claim 3, characterized in that: The shooting module includes an environmental perception camera and an eye-tracking camera; the heart rate detection module uses a PPG sensor installed in the nose pad of the glasses; and the control module includes a central controller and a fourth multi-mode transmission module.
5. The stroke rehabilitation assessment system according to claim 4, characterized in that: Based on a comprehensive rehabilitation assessment score, the stroke rehabilitation assessment system performs automated rehabilitation assessments using sensor data.
6. The stroke rehabilitation assessment system according to claim 5, characterized in that: The comprehensive rehabilitation assessment score includes items such as elbow flexion, wrist circular motion, knee flexion in the supine position, knee extension in the supine position, knee flexion in the sitting position, muscle fatigue level, and emotional regulation ability. Each item is scored as 0, 1, or 2 points. 0 points indicates that the error between the test data and the baseline value is greater than 20%, 1 point indicates that the error between the test data and the baseline value is less than or equal to 20% and greater than 5%, and 2 points indicates that the error between the test data and the baseline value is less than 5%.
7. The stroke rehabilitation assessment system according to claim 6, characterized in that: The stroke rehabilitation assessment system uses an elbow motion detection unit to assess elbow flexion, AR smart glasses to assess wrist circular motion, spherical grasp and emotion regulation ability, a lower limb motion detection unit to assess knee flexion and knee extension, and an electromyography detection unit to assess muscle fatigue.
8. The stroke rehabilitation assessment system according to claim 7, characterized in that: A single assessment process includes initial setup, followed by a health assessment in the order of elbow flexion, wrist circular motion, knee flexion in a supine position, knee extension in a supine position, and knee flexion in a seated position. The elbow motion detection unit, AR smart glasses, and lower limb motion detection unit transmit the scores to the main processor. During the health assessment, the electromyography (EMG) detection unit calculates the score for muscle fatigue, and the AR smart glasses calculate the score for emotional regulation ability. The EMG detection unit and AR smart glasses transmit the calculated scores to the main processor, which then aggregates them to obtain a comprehensive rehabilitation assessment score.
9. The stroke rehabilitation assessment system according to claim 8, characterized in that: The initialization settings specifically include setting the movement patterns and durations for elbow flexion, wrist circular motion, knee flexion in a supine position, knee extension in a supine position, and knee flexion in a seated position. The movement patterns include unilateral movement and bilateral synchronous movement.
10. The stroke rehabilitation assessment system according to any one of claims 1-9, characterized in that: At 11 PM every day, the motion angle detection component encrypts the stored angle information through the first processing chip. Two motion angle detection components communicate with the central processing belt via 5G Wi-Fi and Bluetooth. The motion angle detection components transmit the encrypted angle information to the main storage module of the central processing belt. After the transmission is completed, the motion angle detection components delete the encrypted angle information. The main processor organizes the transmitted data to obtain the data of each evaluation process of the day and re-encrypts the data of each evaluation process.
