Intelligent sensor-based postoperative orthopedic rehabilitation training monitoring device and system

Abstract

This invention discloses a post-orthopedic rehabilitation training monitoring device and system based on intelligent sensing, belonging to the technical field of rehabilitation training monitoring systems. By setting up a data acquisition unit and a data processing unit, it enables the continuous acquisition and unified recording of various types of motion data generated during post-orthopedic rehabilitation training, such as joint angles, muscle activity states, and movement speeds. Real-time monitoring of the patient's training movements is achieved using accelerometers, gyroscopes, and electromyography (EMG) sensors, allowing for the synchronous acquisition of joint movement trajectories, muscle contraction states, and movement rhythms. This enables the system to continuously monitor changes in movement during patient training, providing a complete data foundation for subsequent movement evaluation. Noise removal and outlier processing of the raw data filter out data that may be affected by environmental interference or transient fluctuations during sensor acquisition, thereby reducing the impact of data fluctuations on subsequent analysis.

Description

Technical Field

[0001] This invention relates to the field of rehabilitation training monitoring system technology, specifically to a postoperative rehabilitation training monitoring device and system for orthopedic surgery based on intelligent sensing. Background Technology

[0002] In the intelligent rehabilitation technology system, postoperative rehabilitation training monitoring in orthopedics is an important application area. It primarily monitors and evaluates the rehabilitation training process after orthopedic surgeries such as fracture repair, joint replacement, and ligament reconstruction. Specifically, during postoperative rehabilitation training, patients need to perform joint mobility training, muscle strength training, and motor control training according to a predetermined training plan. Parameters such as changes in joint angles, muscle contraction states, and movement speed during these training processes can be collected in real time by intelligent sensing devices, thus forming the key monitoring objects in the postoperative rehabilitation training monitoring system.

[0003] In current orthopedic postoperative rehabilitation training, patients typically rely on the guidance of doctors or rehabilitation therapists to complete rehabilitation exercises. After leaving the medical institution, the training process is often completed by the patients themselves. Because patients lack real-time monitoring methods during training, it is difficult for them to promptly determine whether their joint range of motion meets training requirements, whether muscle exertion is appropriate, and whether the movement speed matches the training rhythm. Therefore, in the home rehabilitation phase, insufficient range of motion, unstable training rhythm, or inappropriate force application methods are prone to occur.

[0004] In the absence of a real-time feedback mechanism, patients are prone to deviations in movement execution during rehabilitation training. For example, too small a joint range of motion may result in insignificant training effects, or too large a range of motion may cause unnecessary stress on the surgical site. At the same time, uneven muscle exertion or excessively fast movement may also cause impact-induced joint movements, thereby affecting the postoperative recovery process. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an orthopedic postoperative rehabilitation training monitoring device and system based on intelligent sensing, which solves the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solution: a postoperative rehabilitation training monitoring system for orthopedic surgery based on intelligent sensing, including a motion data acquisition and processing module, a motion analysis and evaluation module, a dynamic feedback generation module, an automatic adjustment module for rehabilitation training plans, and a data storage and analysis report module;

[0007] The motion data acquisition and processing module collects the patient's motion data in real time through intelligent sensors, and obtains a monitoring dataset of KW;

[0008] The motion analysis and evaluation module analyzes and evaluates the monitoring dataset KW, obtains the motion evaluation error E, and determines whether the execution of the motion meets the predetermined standard and whether there are any abnormalities.

[0009] The dynamic feedback generation module generates feedback information based on the action assessment error E, provides feedback to the patient in a tactile or visual manner, and obtains the feedback intensity feed.

[0010] The automatic adjustment module of the rehabilitation training plan automatically adjusts the patient's training plan based on the feedback intensity (Feed) and the movement assessment error (E), and obtains the adjusted training time and intensity.

[0011] The data storage and analysis report module collects all patient training data, feedback results, and training adjustment plans, and generates analysis reports.

[0012] Preferably, the motion data acquisition and processing module includes a data acquisition unit and a data processing unit;

[0013] The data acquisition unit collects the patient's motion data in real time through intelligent sensors, including joint angle θ, muscle activity level eg, motion speed v and acceleration a, and fits it into the original dataset YW;

[0014] Among them, the joint angle θ is captured in real time by accelerometer and gyroscope sensors, and the joint angle is calculated by kinematic algorithm;

[0015] The joint angle θ is obtained as follows: the position information of the object in the x-axis and y-axis directions is obtained through sensors, that is, the values ​​of x-coordinate and y-coordinate are obtained; the ratio of y-coordinate and x-coordinate is calculated to obtain their proportional relationship; the arctangent function arctan is used to calculate this ratio; the result is the joint angle θ of the object at that moment.

[0016] Muscle activity level eg is obtained by real-time monitoring of muscle electrical activity through electromyography (EMG) sensors to reflect muscle working intensity and fatigue state. Muscle activity level eg is obtained by taking the absolute value of the difference between the raw EMG signal and the reference EMG signal.

[0017] The velocity v and acceleration a are obtained by acquiring acceleration data in various directions during the motion through an accelerometer and calculating them in combination with the collected timestamp data; the velocity v is obtained by the ratio of the displacement change to the time interval; the acceleration a is obtained by the ratio of the velocity change to the time interval.

[0018] Preferably, the data processing unit cleans and normalizes the collected raw dataset YW to obtain the monitoring dataset KW;

[0019] Cleaning includes noise removal and outlier handling;

[0020] The method for removing noise is as follows: high-frequency noise is removed by using a low-pass filter, while retaining the lower-frequency signal;

[0021] Outlier handling is performed by using a box plot method to detect and remove extreme values; the formula is as follows:

[0022] IQR=Q3-Q1, LB=Q1-1.5×IQR, UB=Q3+1.5×IQR;

[0023] In the formula, IQR represents the interquartile range, specifically the difference between the upper quartile Q3 and the lower quartile Q1. Q1 represents the first quartile, which is the value at the 25th percentile after the dataset is sorted by size, and Q3 represents the third quartile, which is the value at the 75th percentile after the dataset is sorted by size. LB represents the lower bound, used to determine which data points are likely outliers, and UB represents the upper bound, used to determine which data points are likely outliers.

[0024] Data exceeding the upper bound UB and below the lower bound LB will be removed;

[0025] The normalization formula is as follows: Original dataset YW; Monitoring dataset KW;

[0026] KWo=(YWo-minYWo) / (maxYWo-minYWo);

[0027] In the formula, KWo represents the o-th data in the monitoring dataset KW, YWo represents the o-th data in the original dataset YW, minYWo represents the valley value of the o-th data in the original dataset YW, and maxYWo represents the peak value of the o-th data in the original dataset YW.

[0028] Preferably, the motion analysis and evaluation module includes a motion deviation calculation unit and a standard compliance judgment unit;

[0029] The motion deviation calculation unit analyzes the data in the monitoring dataset KW and calculates the joint angle deviation Eθ, muscle activity deviation Eeg, motion velocity deviation Ev, and acceleration deviation Ea.

[0030] The method for obtaining the joint angle deviation Eθ is as follows: the joint angle θ(t) at time t is subtracted from the maximum standard angle θmax, the absolute value is taken, and then divided by the maximum standard angle θmax to obtain the joint angle deviation Eθ.

[0031] The muscle activity deviation Eeg is obtained by subtracting the muscle activity level eg(t) at time t from the maximum electromyographic activity value egmax, taking the absolute value, and then dividing it by the maximum electromyographic activity value egmax to obtain the muscle activity deviation Eeg.

[0032] The motion speed deviation Ev is obtained by subtracting the motion speed v(t) at time t from the maximum motion speed vmax, taking the absolute value, and then dividing it by the maximum motion speed vmax to obtain the motion speed deviation Ev.

[0033] The acceleration deviation Ea is obtained by subtracting the acceleration a(t) at time t from the maximum acceleration amax, taking the absolute value, and then dividing it by the maximum acceleration amax to obtain the acceleration deviation Ea.

[0034] Preferably, the standard compliance judgment unit integrates the acquired joint angle deviation Eθ, muscle activity deviation Eeg, movement speed deviation Ev and acceleration deviation Ea, calculates the motion evaluation error E, and compares it with the preset error threshold TE to determine whether the execution of the motion meets the predetermined standard and whether there is any abnormality.

[0035] The motion evaluation error E is calculated using the following formula:

[0036] E(t)=s1×Eθ(t)+s2×Eeg(t)+s3×Ev(t)+s4×Ea(t);

[0037] In the formula, s1 represents the preset weight value of joint angle deviation, which ranges from 0.2 to 0.4; s2 represents the preset weight value of muscle activity deviation, which ranges from 0.1 to 0.3; s3 represents the preset weight value of motion velocity deviation, which ranges from 0.25 to 0.45; s4 represents the preset weight value of acceleration deviation, which ranges from 0.1 to 0.4; and s1+s2+s3+s4=1. E(t) represents the motion evaluation error at time t.

[0038] The error threshold TE is obtained using the following formula:

[0039] TE=μE(t)+ka×σE(t);

[0040] In the formula, μE(t) represents the mean of the motion evaluation error at time t, σE(t) represents the standard deviation of the motion evaluation error at time t, and ka represents the adjustment coefficient, which takes a value of 0.2 to 0.5.

[0041] The judgment method is as follows:

[0042] When the motion evaluation error E ≤ the error threshold TE, it indicates that the execution meets the standard and there are no abnormalities.

[0043] When the motion evaluation error E > the error threshold TE, it indicates that the execution does not meet the standard and there is an anomaly.

[0044] Preferably, the dynamic feedback generation module includes a feedback intensity calculation unit and a feedback output control unit;

[0045] When there is an abnormality in the action assessment error E, the feedback intensity calculation unit analyzes the patient's current action state, joint angle θ, muscle activity level eg, movement speed v and acceleration a, calculates the degree of action deviation, and generates a feedback intensity value Feed.

[0046] Among them, the degree of movement deviation includes joint angle deviation Dθ, muscle activity deviation Deg, movement rhythm deviation Dv, and movement impact deviation Da;

[0047] The joint angle deviation Dθ is obtained by taking the absolute value of the difference between the current joint angle and the standard training angle.

[0048] The muscle activity deviation (Deg) is obtained by taking the absolute value of the difference between the current electromyographic signal amplitude and the standard electromyographic activity level.

[0049] The deviation of the movement rhythm, Dv, is obtained by taking the absolute value of the difference between the current movement speed and the standard movement speed.

[0050] The motion impact deviation Da is obtained by taking the absolute value of the difference between the current motion acceleration and the standard acceleration.

[0051] The obtained joint angle deviation Dθ, muscle activity deviation Deg, movement rhythm deviation Dv, and movement impact deviation Da are integrated to calculate the overall movement deviation degree MD.

[0052] The method for obtaining the overall movement deviation MD is as follows: square the joint angle deviation Dθ, muscle activity deviation Deg, movement rhythm deviation Dv, and movement impact deviation Da respectively, sum them up, and then take the square root of the summed value to obtain the overall movement deviation MD.

[0053] The feedback intensity value (Feed) is obtained by dividing the overall motion deviation (MD) by 1 and summing it with the absolute value of the motion evaluation error (E).

[0054] Analyze the feedback intensity values ​​(Feed) and classify the feedback levels;

[0055] The feedback levels are categorized as follows:

[0056] When the feedback intensity value Feed < 0.2, it indicates the first feedback level, with no prompt.

[0057] When 0.2 ≤ Feed intensity value ≤ 0.5, it indicates the second feedback level, and a prompt will appear on the screen.

[0058] When 0.5 < Feed intensity value Feed ≤ 1, it indicates the third feedback level, which includes visual and vibration cues;

[0059] When 1 < Feed intensity value Feed, it indicates the fourth feedback level, a strong vibration warning.

[0060] Preferably, the feedback output control unit generates specific feedback signals based on the feedback intensity Feed, including the feedback signal amplitude sf, the feedback vibration intensity Vb, and the visual cue level Lv; and controls the feedback device to output cue information so that the patient can adjust their movements in a timely manner.

[0061] The feedback signal amplitude sf is obtained as follows: Sf = ln(1 + Feed); where ln represents the logarithmic function;

[0062] The feedback vibration intensity Vb is obtained as follows: Vb = Vmax × Sf; where Vmax represents the maximum vibration intensity of the equipment.

[0063] The visual cue level Lv is obtained as follows: Lv = ⌊3 × Sf⌋; where ⌊⌋ represents the floor function.

[0064] Get different screen cues based on the visual cue level (Lv):

[0065] When the visual cue level Lv=0, there is no cue;

[0066] When the visual cue level Lv=1, a yellow warning is displayed on the screen.

[0067] When the visual cue level Lv=2, the screen displays an orange cue.

[0068] When the visual cue level Lv=3, a red cue is displayed on the screen.

[0069] Preferably, the automatic adjustment module for the rehabilitation training plan includes a training status assessment unit and a training plan adjustment unit;

[0070] The training status assessment unit calculates the patient's current training status index SQ based on the feedback intensity Feed and the movement assessment error E, and determines whether the patient's current training is in a stable state, a deviation state, or an insufficient load state.

[0071] The training state index SQ is obtained by taking the square root of the sum of the square of the feedback intensity Feed and the square of the action evaluation error E.

[0072] When the training state index SQ ≤ 0.7, it indicates that the training action is stable;

[0073] When the training state index SQ > 0.7, it indicates that the action deviates and the training action is abnormal;

[0074] The motion stability KQ is further calculated based on the training state index SQ.

[0075] Action stability KQ is obtained by dividing 1 by 1 and summing it with the training state index SQ. When the action is stable, the training state index SQ is small, and the action stability KQ is close to 1; when the action deviates significantly, the training state index SQ increases, and the action stability KQ gradually decreases.

[0076] The training plan adjustment unit automatically adjusts the training time and intensity based on the training state index SQ and the movement stability KQ, obtains the adjusted training time and intensity, and generates a new training plan.

[0077] The adjusted training time is obtained by multiplying the current training time by the movement stability KQ; when the movement stability is high, the training time remains close to the original plan; when the movement stability decreases, the training time is automatically reduced to make the training pace more moderate.

[0078] The adjusted training intensity is obtained by dividing the current training intensity by 1 and summing it with the training state index SQ; when the training load level is high, the training intensity is reduced; when the load level is low, the training intensity is close to the original plan.

[0079] Based on the adjusted training time and intensity, calculate the training rhythm correction amount R and revise the training plan accordingly.

[0080] The training pacing adjustment R is obtained by multiplying the current training time and the current training intensity by the product of the adjusted training time and the adjusted training intensity.

[0081] When the training pace correction R < 0.5, it indicates that the training load is reduced and the rest time is increased.

[0082] When 0.5 ≤ training pace correction R ≤ 1.2, it means that the training plan will remain in its current adjusted state;

[0083] When 1.2 < training pace correction R, the training state is stable, and the training task is gradually increased in subsequent training cycles.

[0084] Preferably, the data storage and analysis report module includes a data encapsulation and cloud storage unit and a rehabilitation progress analysis and report generation unit;

[0085] The data encapsulation and cloud storage unit receives the generated training data, including the monitoring dataset KW, action evaluation error E, feedback intensity Feed, and training plan adjustment results;

[0086] First, all types of data are sorted according to timestamps and a training time series is established. Then, the data is divided into several training segments based on continuous action changes, and the corresponding action error, feedback triggering situation, and training plan adjustment information are recorded in each segment.

[0087] Next, a data index is created for each training segment to form a training record structure that includes segment number, time range, action parameters, and feedback events;

[0088] Finally, the packaged training records are uploaded to the cloud database and stored in layers according to the original data layer, training segment layer and event index layer.

[0089] The rehabilitation progress analysis and report generation unit aggregates training segments based on training dates or training cycles, and calculates the training duration, number of movement segments, and number of feedback triggers within each cycle.

[0090] Subsequently, the joint angle data in each segment were statistically processed to obtain the maximum and minimum angles of each segment, and the range of motion of the joints was obtained accordingly. At the same time, segment-level statistics were performed on the movement assessment error sequence to count the duration and frequency of errors. Then, combined with feedback intensity and training plan adjustment records, the deviations that occurred during training and the corresponding adjustment behaviors were analyzed. Finally, a rehabilitation progress report was generated based on changes in range of motion, error changes, and training adjustments. The report included range of motion, training stability, recovery progress, and subsequent training suggestions.

[0091] The intelligent sensing-based post-orthopedic rehabilitation training monitoring device includes an inertial sensor module, a surface electromyography acquisition module, a flexible angle sensor, a vibration feedback module, and a mobile data terminal.

[0092] This invention provides a postoperative rehabilitation training monitoring device and system based on intelligent sensing in orthopedics, which has the following beneficial effects:

[0093] (1) During system operation, by setting up data acquisition and data processing units, various types of motion data, such as joint angles, muscle activity status, and movement speed, generated during orthopedic postoperative rehabilitation training can be continuously collected and formed into a unified data record. Accelerometers, gyroscopes, and electromyography sensors are used to monitor the patient's training movements in real time, so that joint movement trajectories, muscle contraction status, and movement rhythm can be acquired synchronously. This allows the system to continuously monitor the changes in movement during the patient's training process, providing a complete data foundation for subsequent movement assessment.

[0094] By removing noise and processing outliers from the raw data, data that may be affected by environmental interference or transient fluctuations during sensor acquisition is filtered out, thus reducing the impact of data fluctuations on subsequent analysis. Simultaneously, by detecting and removing outliers, extreme data deviating from the normal range are excluded from subsequent calculations, making the monitoring data more stable. Data normalization allows different types of data to be represented on a unified scale, enabling standardized analysis of data from different sources, such as joint angles, muscle activity states, and movement speeds.

[0095] (2) By setting up a movement deviation calculation unit and a standard compliance judgment unit, the system can comprehensively analyze the multidimensional movement data generated during the patient's rehabilitation training. By calculating the deviations in joint angle, muscle activity state, movement speed, and acceleration respectively, the system can reflect the differences between the patient's current training movement and the preset training standard from multiple perspectives. Compared with the method of judging based on only a single parameter, this structure can simultaneously reflect the changes in joint range of motion, muscle contraction state, and movement rhythm, making the assessment of movement state more comprehensive and thus more accurately identifying deviations in training movements.

[0096] (3) By comprehensively analyzing the degree of movement deviation and the error in movement assessment, and dividing the calculated feedback intensity into multiple feedback levels, the system can output different forms of prompting information according to different degrees of movement deviation. When the degree of movement deviation is small, the system does not generate prompts. However, when the degree of movement deviation gradually increases, the system reminds the patient to adjust the training movements in sequence through screen prompts, visual and vibration prompts, and strong vibration prompts. This allows the feedback prompts to gradually increase with the change in the degree of movement deviation, so that the patient can promptly know the current movement status and make adjustments during the training process.

[0097] (4) By setting up a training status assessment unit and a training plan adjustment unit, the system can comprehensively judge the patient's current training status based on feedback intensity and movement assessment error, and dynamically adjust the training plan accordingly. By continuously assessing the patient's movement status during training, the system can determine whether the current training movement is stable or deviates from its target state, thus allowing the training plan to be adjusted accordingly as the patient's movement status changes. Compared with a fixed training plan, this structure can adjust the training time and intensity based on the patient's actual movement performance during training, ensuring that the training load corresponds to the patient's current physical state. Attached Figure Description

[0098] Figure 1 This is a schematic diagram of the block flow of the intelligent sensing-based post-orthopedic rehabilitation training monitoring system of the present invention.

[0099] Figure 2 This is a schematic diagram of the visual cue level acquisition process of the present invention;

[0100] Figure 3 This is a schematic diagram of the training state index generation process of the present invention;

[0101] Figure 4 This is a schematic diagram of the training plan generation process of the present invention. Detailed Implementation

[0102] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0103] Example 1

[0104] This invention provides an intelligent sensing-based post-orthopedic rehabilitation training monitoring system. Please refer to [link / reference]. Figures 1 to 4 It includes a motion data acquisition and processing module, a motion analysis and evaluation module, a dynamic feedback generation module, an automatic adjustment module for rehabilitation training plans, and a data storage and analysis report module;

[0105] The motion data acquisition and processing module collects the patient's motion data in real time through intelligent sensors, and obtains a monitoring dataset of KW;

[0106] The motion analysis and evaluation module analyzes and evaluates the monitoring dataset KW, obtains the motion evaluation error E, and determines whether the execution of the motion meets the predetermined standard and whether there are any abnormalities.

[0107] The dynamic feedback generation module generates feedback information based on the action assessment error E, provides feedback to the patient in a tactile or visual manner, and obtains the feedback intensity feed.

[0108] The automatic adjustment module of the rehabilitation training plan automatically adjusts the patient's training plan based on the feedback intensity (Feed) and the movement assessment error (E), and obtains the adjusted training time and intensity.

[0109] The data storage and analysis report module collects all patient training data, feedback results, and training adjustment plans, and generates analysis reports.

[0110] In this embodiment, by constructing a motion data acquisition and processing module, a motion analysis and evaluation module, a dynamic feedback generation module, an automatic adjustment module for rehabilitation training plans, and a data storage and analysis report module, data such as joint angles, muscle activity states, and movement speed generated by patients during rehabilitation training can be continuously collected and subjected to structured analysis. By monitoring the training movement state in real time and calculating the movement evaluation error, the system can identify deviations between the training movements and predetermined training standards, thereby enabling patients to obtain objective movement state information during training.

[0111] The dynamic feedback generation module converts motion assessment results into tactile or visual cues, allowing patients to promptly identify whether their current movement deviates from the predetermined range during training. This enables them to proactively adjust joint range of motion, movement rhythm, and muscle exertion during training, ensuring the exercise remains within the preset training range. Compared to methods that only review data after training, this structure allows training feedback to directly impact the training process itself, fostering a more stable movement execution pattern in patients.

[0112] The automatic adjustment module of the rehabilitation training plan dynamically adjusts the training time and intensity based on feedback intensity and movement assessment errors, allowing the training plan to change according to the patient's movement performance. When the system detects significant deviations in training movements, the training plan can reduce the training load or increase the interval time; when training movements remain stable, the training plan can maintain or gradually expand the training tasks, ensuring that the rehabilitation training process corresponds to the patient's current physical condition.

[0113] Furthermore, the data storage and analysis reporting module uniformly records training data, feedback information, and training adjustment results, generating a rehabilitation progress report. This allows doctors to comprehensively assess the patient's recovery status based on changes in movement across multiple training cycles and adjust subsequent rehabilitation plans accordingly. This structure enables patients to receive continuous movement monitoring, process feedback, and training plan adjustments during rehabilitation training, thus making the training process more standardized and reducing unstable training due to deviations in movement execution.

[0114] Example 2

[0115] This embodiment is an explanation based on Embodiment 1. Please refer to it. Figure 1 Specifically: the motion data acquisition and processing module includes a data acquisition unit and a data processing unit;

[0116] The data acquisition unit collects the patient's motion data in real time through intelligent sensors, including joint angle θ, muscle activity level eg, motion speed v and acceleration a, and fits it into the original dataset YW;

[0117] Among them, the joint angle θ is captured in real time by accelerometer and gyroscope sensors, and the joint angle is calculated by kinematic algorithm;

[0118] The joint angle θ is obtained as follows: the position information of the object in the x-axis and y-axis directions is obtained through sensors, that is, the values ​​of x-coordinate and y-coordinate are obtained; the ratio of y-coordinate and x-coordinate is calculated to obtain their proportional relationship; the arctangent function arctan is used to calculate this ratio; and the joint angle θ is obtained.

[0119] Muscle activity level, e.g., is monitored in real time by electromyography (EMG) sensors; muscle activity level, e.g., is obtained by taking the absolute value of the difference between the raw EMG signal and the reference EMG signal.

[0120] The velocity v and acceleration a are obtained by acquiring acceleration data in various directions during the motion through an accelerometer and calculating them in combination with the collected timestamp data; the velocity v is obtained by the ratio of the displacement change to the time interval; the acceleration a is obtained by the ratio of the velocity change to the time interval.

[0121] The data processing unit cleans and normalizes the collected raw dataset YW to obtain the monitoring dataset KW;

[0122] Cleaning includes noise removal and outlier handling;

[0123] The method for removing noise is as follows: high-frequency noise is removed by using a low-pass filter, while retaining the lower-frequency signal;

[0124] Outlier handling is performed by using a box plot method to detect and remove extreme values; the formula is as follows:

[0125] IQR=Q3-Q1, LB=Q1-1.5×IQR, UB=Q3+1.5×IQR;

[0126] In the formula, IQR represents the interquartile range, specifically the difference between the upper quartile Q3 and the lower quartile Q1. Q1 represents the first quartile, which is the value at the 25th percentile after the dataset is sorted by size. Q3 represents the third quartile, which is the value at the 75th percentile after the dataset is sorted by size. LB represents the lower bound, and UB represents the upper bound.

[0127] Data exceeding the upper bound UB and below the lower bound LB will be removed;

[0128] The normalization formula is as follows: Original dataset YW; Monitoring dataset KW;

[0129] KWo=(YWo-minYWo) / (maxYWo-minYWo);

[0130] In the formula, KWo represents the o-th data in the monitoring dataset KW, YWo represents the o-th data in the original dataset YW, minYWo represents the valley value of the o-th data in the original dataset YW, and maxYWo represents the peak value of the o-th data in the original dataset YW.

[0131] In this embodiment, by setting up a data acquisition unit and a data processing unit, various types of motion data, such as joint angles, muscle activity states, and movement speed, generated during orthopedic postoperative rehabilitation training can be continuously collected and formed into a unified data record. Accelerometers, gyroscopes, and electromyography sensors are used to monitor the patient's training movements in real time, enabling the synchronous acquisition of joint movement trajectories, muscle contraction states, and movement rhythms. This allows the system to continuously monitor changes in movement during patient training, providing a complete data foundation for subsequent movement evaluation.

[0132] By removing noise and processing outliers from the raw data, data that may be affected by environmental interference or transient fluctuations during sensor acquisition is filtered out, thus reducing the impact of data fluctuations on subsequent analysis. Simultaneously, by detecting and removing outliers, extreme data deviating from the normal range are excluded from subsequent calculations, making the monitoring data more stable. Data normalization allows different types of data to be represented on a unified scale, enabling standardized analysis of data from different sources, such as joint angles, muscle activity states, and movement speeds.

[0133] This structure enables the processing and standardization of multi-source motion data generated during orthopedic postoperative rehabilitation training before it enters the movement assessment stage, ensuring the continuity and comparability of the monitoring data. Compared to methods relying on a single motion parameter for monitoring, this structure simultaneously reflects joint movement and muscle activity, allowing the system to more comprehensively reflect changes in the patient's training movements. This provides a more stable data foundation for subsequent movement analysis and feedback, enabling continuous monitoring of movement status during rehabilitation training.

[0134] Example 3

[0135] This embodiment is an explanation based on Embodiment 2. Please refer to it. Figure 1 Specifically: the motion analysis and evaluation module includes a motion deviation calculation unit and a standard compliance judgment unit;

[0136] The motion deviation calculation unit analyzes the data in the monitoring dataset KW and calculates the joint angle deviation Eθ, muscle activity deviation Eeg, motion velocity deviation Ev, and acceleration deviation Ea.

[0137] The method for obtaining the joint angle deviation Eθ is as follows: the joint angle θ(t) at time t is subtracted from the maximum standard angle θmax, the absolute value is taken, and then divided by the maximum standard angle θmax to obtain the joint angle deviation Eθ.

[0138] The muscle activity deviation Eeg is obtained by subtracting the muscle activity level eg(t) at time t from the maximum electromyographic activity value egmax, taking the absolute value, and then dividing it by the maximum electromyographic activity value egmax to obtain the muscle activity deviation Eeg.

[0139] The motion speed deviation Ev is obtained by subtracting the motion speed v(t) at time t from the maximum motion speed vmax, taking the absolute value, and then dividing it by the maximum motion speed vmax to obtain the motion speed deviation Ev.

[0140] The acceleration deviation Ea is obtained by subtracting the acceleration a(t) at time t from the maximum acceleration amax, taking the absolute value, and then dividing it by the maximum acceleration amax to obtain the acceleration deviation Ea.

[0141] The standard compliance judgment unit integrates the acquired joint angle deviation Eθ, muscle activity deviation Eeg, movement speed deviation Ev and acceleration deviation Ea, calculates the motion evaluation error E, and compares it with the preset error threshold TE to determine whether the execution of the motion meets the predetermined standard and whether there are any abnormalities.

[0142] The motion evaluation error E is calculated using the following formula:

[0143] E(t)=s1×Eθ(t)+s2×Eeg(t)+s3×Ev(t)+s4×Ea(t);

[0144] In the formula, s1 represents the preset weight value of joint angle deviation, s2 represents the preset weight value of muscle activity deviation, s3 represents the preset weight value of movement speed deviation, s4 represents the preset weight value of acceleration deviation, and s1+s2+s3+s4=1, and E(t) represents the motion evaluation error at time t.

[0145] The error threshold TE is obtained using the following formula:

[0146] TE=μE(t)+ka×σE(t);

[0147] In the formula, μE(t) represents the mean of the motion evaluation error at time t, σE(t) represents the standard deviation of the motion evaluation error at time t, and ka represents the adjustment coefficient, which takes a value of 0.2 to 0.5.

[0148] The judgment method is as follows:

[0149] When the motion evaluation error E ≤ the error threshold TE, it indicates that the execution meets the standard and there are no abnormalities.

[0150] When the motion evaluation error E > the error threshold TE, it indicates that the execution does not meet the standard and there is an anomaly.

[0151] In this embodiment, by setting up a movement deviation calculation unit and a standard compliance judgment unit, the system can comprehensively analyze multidimensional movement data generated during the patient's rehabilitation training. By calculating the deviations in joint angles, muscle activity states, movement speed, and acceleration respectively, the system can reflect the differences between the patient's current training movements and preset training standards from multiple perspectives. Compared with methods that rely on only a single parameter for judgment, this structure can simultaneously reflect changes in joint range of motion, muscle contraction state, and movement rhythm, making the assessment of movement state more comprehensive and thus enabling more accurate identification of deviations in training movements.

[0152] By integrating various deviation data and forming a motion assessment error, the system can uniformly evaluate the overall state of the patient's current training movements and determine whether the training movements are within the normal range by comparing them with preset error thresholds. By analyzing the statistical changes in motion assessment errors, the system can dynamically judge the stability of movements based on error changes during training, thus enabling the motion assessment results to reflect the actual changes in the patient's training process, rather than relying solely on fixed thresholds.

[0153] This structure allows for continuous monitoring and analysis of changes in patient movements during rehabilitation training. When deviations from the prescribed movements occur, the system can promptly identify abnormalities, providing a basis for subsequent feedback and adjustments to the training plan. Compared to methods relying solely on manual observation or post-training data review, this structure continuously assesses movement execution during training, enabling timely identification of whether movements meet predetermined training requirements. This leads to a more stable rehabilitation training process and reduces non-standard training caused by movement deviations.

[0154] Example 4

[0155] This embodiment is an explanation based on Embodiment 3. Please refer to it. Figure 2 Specifically: the dynamic feedback generation module includes a feedback intensity calculation unit and a feedback output control unit;

[0156] When there is an abnormality in the action assessment error E, the feedback intensity calculation unit analyzes the patient's current action state, joint angle θ, muscle activity level eg, movement speed v and acceleration a, calculates the degree of action deviation, and generates a feedback intensity value Feed.

[0157] Among them, the degree of movement deviation includes joint angle deviation Dθ, muscle activity deviation Deg, movement rhythm deviation Dv, and movement impact deviation Da;

[0158] The joint angle deviation Dθ is obtained by taking the absolute value of the difference between the current joint angle and the standard training angle.

[0159] The muscle activity deviation (Deg) is obtained by taking the absolute value of the difference between the current electromyographic signal amplitude and the standard electromyographic activity level.

[0160] The deviation of the movement rhythm, Dv, is obtained by taking the absolute value of the difference between the current movement speed and the standard movement speed.

[0161] The motion impact deviation Da is obtained by taking the absolute value of the difference between the current motion acceleration and the standard acceleration.

[0162] The obtained joint angle deviation Dθ, muscle activity deviation Deg, movement rhythm deviation Dv, and movement impact deviation Da are integrated to calculate the overall movement deviation degree MD.

[0163] The method for obtaining the overall movement deviation MD is as follows: square the joint angle deviation Dθ, muscle activity deviation Deg, movement rhythm deviation Dv, and movement impact deviation Da respectively, sum them up, and then take the square root of the summed value to obtain the overall movement deviation MD.

[0164] The feedback intensity value (Feed) is obtained by dividing the overall motion deviation (MD) by 1 and summing it with the absolute value of the motion evaluation error (E).

[0165] Analyze the feedback intensity values ​​(Feed) and classify the feedback levels;

[0166] The feedback levels are categorized as follows:

[0167] When the feedback intensity value Feed < 0.2, it indicates the first feedback level, with no prompt.

[0168] When 0.2 ≤ Feed intensity value ≤ 0.5, it indicates the second feedback level, and a prompt will appear on the screen.

[0169] When 0.5 < Feed intensity value Feed ≤ 1, it indicates the third feedback level, which includes visual and vibration cues;

[0170] When 1 < Feed intensity value Feed, it indicates the fourth feedback level, a strong vibration warning.

[0171] The feedback output control unit generates specific feedback signals based on the feedback intensity Feed, including the feedback signal amplitude sf, the feedback vibration intensity Vb, and the visual cue level Lv;

[0172] The feedback signal amplitude sf is obtained as follows: Sf = ln(1 + Feed); where ln represents the logarithmic function;

[0173] The feedback vibration intensity Vb is obtained as follows: Vb = Vmax × Sf; where Vmax represents the maximum vibration intensity of the equipment.

[0174] The visual cue level Lv is obtained as follows: Lv = ⌊3 × Sf⌋; where ⌊⌋ represents the floor function.

[0175] Get different screen cues based on the visual cue level (Lv):

[0176] When the visual cue level Lv=0, there is no cue;

[0177] When the visual cue level Lv=1, a yellow warning is displayed on the screen.

[0178] When the visual cue level Lv=2, the screen displays an orange cue.

[0179] When the visual cue level Lv=3, a red cue is displayed on the screen.

[0180] In this embodiment, by setting up a feedback intensity calculation unit and a feedback output control unit, the system can further analyze the patient's current movement state when abnormalities are detected in the training movement. It calculates the degree of movement deviation based on multiple motion parameters such as joint angles, muscle activity, movement speed, and acceleration, thereby generating a feedback intensity corresponding to the current movement state. By simultaneously considering factors such as movement amplitude, muscle exertion state, and movement rhythm, the system can more comprehensively reflect the deviation of the patient's training movement, ensuring that the feedback information corresponds to the patient's current movement state.

[0181] By comprehensively analyzing the degree of movement deviation and movement assessment error, and dividing the calculated feedback intensity into multiple feedback levels, the system can output different forms of prompts based on the degree of movement deviation. When the degree of movement deviation is small, the system does not generate prompts. However, as the degree of movement deviation gradually increases, the system sequentially reminds the patient to adjust their training movements through screen prompts, visual and vibration prompts, and strong vibration prompts. This allows the feedback prompts to gradually increase with the change in the degree of movement deviation, enabling the patient to promptly understand their current movement status and make adjustments during training.

[0182] The feedback output control unit converts feedback intensity into vibration intensity and visual cue levels, allowing tactile and visual cues to change accordingly based on movement deviations. This enables patients to perceive their movement execution through intuitive prompts during training. Compared to providing data for review only after training, this structure allows feedback information to directly impact the training process. Patients can adjust joint range of motion, movement rhythm, and muscle exertion during movement, ensuring the training movements remain within the preset range. This reduces training irregularities caused by movement deviations and helps maintain the stability of the rehabilitation training process.

[0183] Example 5

[0184] This embodiment is an explanation based on Embodiment 4. Please refer to it. Figure 3 and Figure 4 Specifically: the automatic adjustment module for rehabilitation training plans includes a training status assessment unit and a training plan adjustment unit;

[0185] The training status assessment unit calculates the patient's current training status index SQ based on the feedback intensity Feed and the movement assessment error E, and determines whether the patient's current training is in a stable state, a deviation state, or an insufficient load state.

[0186] The training state index SQ is obtained by taking the square root of the sum of the square of the feedback intensity Feed and the square of the action evaluation error E.

[0187] When the training state index SQ ≤ 0.7, it indicates that the training action is stable;

[0188] When the training state index SQ > 0.7, it indicates that the action deviates and the training action is abnormal;

[0189] The motion stability KQ is further calculated based on the training state index SQ.

[0190] Action stability KQ is obtained by dividing 1 by 1 and summing it with the training state index SQ;

[0191] The training plan adjustment unit automatically adjusts the training time and intensity based on the training state index SQ and the movement stability KQ, obtains the adjusted training time and intensity, and generates a new training plan.

[0192] The adjusted training time is obtained by multiplying the current training time by the movement stability KQ;

[0193] The adjusted training intensity is obtained by dividing the current training intensity by 1 and summing it with the training state index SQ;

[0194] Based on the adjusted training time and intensity, calculate the training rhythm correction amount R and revise the training plan accordingly.

[0195] The training pacing adjustment R is obtained by multiplying the current training time and the current training intensity by the product of the adjusted training time and the adjusted training intensity.

[0196] When the training pace correction R < 0.5, it indicates that the training load is reduced and the rest time is increased.

[0197] When 0.5 ≤ training pace correction R ≤ 1.2, it means that the training plan will remain in its current adjusted state;

[0198] When 1.2 < training pace correction R, the training state is stable, and the training task is gradually increased in subsequent training cycles.

[0199] The data storage and analysis report module includes a data encapsulation and cloud storage unit and a rehabilitation progress analysis and report generation unit;

[0200] The data encapsulation and cloud storage unit receives the generated training data, including the monitoring dataset KW, action evaluation error E, feedback intensity Feed, and training plan adjustment results;

[0201] First, all types of data are sorted according to timestamps and a training time series is established. Then, the data is divided into several training segments based on continuous action changes, and the corresponding action error, feedback triggering situation, and training plan adjustment information are recorded in each segment.

[0202] Next, a data index is created for each training segment to form a training record structure that includes segment number, time range, action parameters, and feedback events;

[0203] Finally, the packaged training records are uploaded to the cloud database and stored in layers according to the original data layer, training segment layer and event index layer.

[0204] The rehabilitation progress analysis and report generation unit aggregates training segments based on training dates or training cycles, and calculates the training duration, number of movement segments, and number of feedback triggers within each cycle.

[0205] Subsequently, the joint angle data in each segment were statistically processed to obtain the maximum and minimum angles of each segment, and the range of motion of the joints was obtained accordingly. At the same time, segment-level statistics were performed on the movement assessment error sequence to count the duration and frequency of errors. Then, combined with feedback intensity and training plan adjustment records, the deviations that occurred during training and the corresponding adjustment behaviors were analyzed. Finally, a rehabilitation progress report was generated based on changes in range of motion, error changes, and training adjustments. The report included range of motion, training stability, recovery progress, and subsequent training suggestions.

[0206] In this embodiment, by setting up a training status assessment unit and a training plan adjustment unit, the system can comprehensively judge the patient's current training status based on feedback intensity and movement assessment error, and dynamically adjust the training plan accordingly. By continuously evaluating the patient's movement status during training, the system can determine whether the current training movement is stable or deviates from its intended state, thus allowing the training plan to be adjusted accordingly to changes in the patient's movement status. Compared to a fixed training plan, this structure can adjust the training time and intensity based on the patient's actual movement performance during training, ensuring that the training load corresponds to the patient's current physical state.

[0207] By synchronously adjusting training time and intensity, and further refining the training plan in conjunction with training rhythm correction, the system can continuously control the training rhythm during the training process. When the system detects significant deviations in training movements, it can reduce the training load by lowering the training intensity or increasing rest time. When the training movements remain stable, it can gradually increase the training tasks in subsequent training cycles, allowing the rehabilitation training process to gradually change with the patient's recovery, thus making the training process smoother and reducing discomfort caused by training load mismatch.

[0208] By setting up data encapsulation and cloud storage units, as well as rehabilitation progress analysis and report generation units, the system enables the unified recording and periodic analysis of patients' movement data, feedback information, and training adjustment results during training. Through statistical analysis of training segments and changes in joint range of motion, movement errors, and training adjustments, the system generates rehabilitation progress reports reflecting the patient's recovery status. This allows doctors to assess the patient's recovery based on data changes across multiple training cycles and adjust subsequent rehabilitation plans accordingly.

[0209] Example 6

[0210] For a smart sensor-based post-orthopedic rehabilitation training monitoring device, please refer to [link / reference]. Figure 1 Specifically, it includes an inertial sensor module, a surface electromyography acquisition module, a flexible angle sensor, a vibration feedback module, and a mobile data terminal;

[0211] In this embodiment, by incorporating an inertial sensor module, a surface electromyography (EMG) acquisition module, a flexible angle sensor, a vibration feedback module, and a mobile data terminal, information such as joint movement status, muscle activity status, and movement rhythm involved in post-orthopedic rehabilitation training can be simultaneously collected and recorded. The inertial sensor module and flexible angle sensor detect joint movement trajectories and angles, allowing continuous monitoring of changes in the patient's range of motion during training. Simultaneously, the EMG acquisition module collects data on muscle contraction states, enabling the system to acquire information on muscle exertion during training, thus forming a relatively complete source of training movement data.

[0212] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A postoperative rehabilitation training monitoring system for orthopedic surgery based on intelligent sensing, characterized in that: It includes a motion data acquisition and processing module, a motion analysis and evaluation module, a dynamic feedback generation module, an automatic adjustment module for rehabilitation training plans, and a data storage and analysis report module. The motion data acquisition and processing module collects the patient's motion data in real time through intelligent sensors, and obtains a monitoring dataset of KW; The motion analysis and evaluation module analyzes and evaluates the monitoring dataset KW, obtains the motion evaluation error E, and determines whether the execution of the motion meets the predetermined standard and whether there are any abnormalities. The dynamic feedback generation module generates feedback information based on the action assessment error E, provides feedback to the patient in a tactile or visual manner, and obtains the feedback intensity feed. The automatic adjustment module of the rehabilitation training plan automatically adjusts the patient's training plan based on the feedback intensity (Feed) and the movement assessment error (E), and obtains the adjusted training time and intensity. The data storage and analysis report module collects all patient training data, feedback results, and training adjustment plans, and generates analysis reports.

2. The orthopedic postoperative rehabilitation training monitoring system based on intelligent sensing according to claim 1, characterized in that: The motion data acquisition and processing module includes a data acquisition unit and a data processing unit; The data acquisition unit collects the patient's motion data in real time through intelligent sensors, including joint angle θ, muscle activity level eg, motion speed v and acceleration a, and fits it into the original dataset YW; Among them, the joint angle θ is captured in real time by accelerometer and gyroscope sensors, and the joint angle is calculated by kinematic algorithm; The joint angle θ is obtained as follows: the position information of the object in the x-axis and y-axis directions is obtained through sensors, that is, the values ​​of x-coordinate and y-coordinate are obtained; the ratio of y-coordinate and x-coordinate is calculated to obtain their proportional relationship; the arctangent function arctan is used to calculate this ratio; and the joint angle θ is obtained. Muscle activity level, e.g., is monitored in real time by electromyography (EMG) sensors; muscle activity level, e.g., is obtained by taking the absolute value of the difference between the raw EMG signal and the reference EMG signal. The velocity v and acceleration a are obtained by acquiring acceleration data in various directions during the motion through an accelerometer and calculating them in combination with the collected timestamp data; the velocity v is obtained by the ratio of the displacement change to the time interval; the acceleration a is obtained by the ratio of the velocity change to the time interval.

3. The orthopedic postoperative rehabilitation training monitoring system based on intelligent sensing according to claim 2, characterized in that: The data processing unit cleans and normalizes the collected raw dataset YW to obtain the monitoring dataset KW; Cleaning includes noise removal and outlier handling; The method for removing noise is as follows: high-frequency noise is removed by using a low-pass filter, while retaining the lower-frequency signal; Outlier handling is performed by using a box plot method to detect and remove extreme values; the formula is as follows: IQR=Q3-Q1, LB=Q1-1.5×IQR, UB=Q3+1.5×IQR; In the formula, IQR represents the interquartile range, specifically the difference between the upper quartile Q3 and the lower quartile Q1. Q1 represents the first quartile, which is the value at the 25th percentile after the dataset is sorted by size. Q3 represents the third quartile, which is the value at the 75th percentile after the dataset is sorted by size. LB represents the lower bound, and UB represents the upper bound. Data exceeding the upper bound UB and below the lower bound LB will be removed; The normalization formula is as follows: Original dataset YW; Monitoring dataset KW; KWo=(YWo-minYWo) / (maxYWo-minYWo); In the formula, KWo represents the o-th data in the monitoring dataset KW, YWo represents the o-th data in the original dataset YW, minYWo represents the valley value of the o-th data in the original dataset YW, and maxYWo represents the peak value of the o-th data in the original dataset YW.

4. The postoperative rehabilitation training monitoring system for orthopedic surgery based on intelligent sensing according to claim 3, characterized in that: The motion analysis and evaluation module includes a motion deviation calculation unit and a standard compliance judgment unit; The motion deviation calculation unit analyzes the data in the monitoring dataset KW and calculates the joint angle deviation Eθ, muscle activity deviation Eeg, motion velocity deviation Ev, and acceleration deviation Ea. The method for obtaining the joint angle deviation Eθ is as follows: the joint angle θ(t) at time t is subtracted from the maximum standard angle θmax, the absolute value is taken, and then divided by the maximum standard angle θmax to obtain the joint angle deviation Eθ. The muscle activity deviation Eeg is obtained by subtracting the muscle activity level eg(t) at time t from the maximum electromyographic activity value egmax, taking the absolute value, and then dividing it by the maximum electromyographic activity value egmax to obtain the muscle activity deviation Eeg. The motion speed deviation Ev is obtained by subtracting the motion speed v(t) at time t from the maximum motion speed vmax, taking the absolute value, and then dividing it by the maximum motion speed vmax to obtain the motion speed deviation Ev. The acceleration deviation Ea is obtained by subtracting the acceleration a(t) at time t from the maximum acceleration amax, taking the absolute value, and then dividing it by the maximum acceleration amax to obtain the acceleration deviation Ea.

5. The orthopedic postoperative rehabilitation training monitoring system based on intelligent sensing according to claim 4, characterized in that: The standard compliance judgment unit integrates the acquired joint angle deviation Eθ, muscle activity deviation Eeg, movement speed deviation Ev and acceleration deviation Ea, calculates the motion evaluation error E, and compares it with the preset error threshold TE to determine whether the execution of the motion meets the predetermined standard and whether there are any abnormalities. The motion evaluation error E is calculated using the following formula: E(t)=s1×Eθ(t)+s2×Eeg(t)+s3×Ev(t)+s4×Ea(t); In the formula, s1 represents the preset weight value of joint angle deviation, s2 represents the preset weight value of muscle activity deviation, s3 represents the preset weight value of movement speed deviation, s4 represents the preset weight value of acceleration deviation, and s1+s2+s3+s4=1, and E(t) represents the motion evaluation error at time t. The error threshold TE is obtained using the following formula: TE=μE(t)+ka×σE(t); In the formula, μE(t) represents the mean of the motion evaluation error at time t, σE(t) represents the standard deviation of the motion evaluation error at time t, and ka represents the adjustment coefficient, which takes a value of 0.2 to 0.

5. The judgment method is as follows: When the motion evaluation error E ≤ the error threshold TE, it indicates that the execution meets the standard and there are no abnormalities. When the motion evaluation error E > the error threshold TE, it indicates that the execution does not meet the standard and there is an anomaly.

6. The orthopedic postoperative rehabilitation training monitoring system based on intelligent sensing according to claim 5, characterized in that: The dynamic feedback generation module includes a feedback intensity calculation unit and a feedback output control unit; When there is an abnormality in the action assessment error E, the feedback intensity calculation unit analyzes the patient's current action state, joint angle θ, muscle activity level eg, movement speed v and acceleration a, calculates the degree of action deviation, and generates a feedback intensity value Feed. Among them, the degree of movement deviation includes joint angle deviation Dθ, muscle activity deviation Deg, movement rhythm deviation Dv, and movement impact deviation Da; The joint angle deviation Dθ is obtained by taking the absolute value of the difference between the current joint angle and the standard training angle. The muscle activity deviation (Deg) is obtained by taking the absolute value of the difference between the current electromyographic signal amplitude and the standard electromyographic activity level. The deviation of the movement rhythm, Dv, is obtained by taking the absolute value of the difference between the current movement speed and the standard movement speed. The motion impact deviation Da is obtained by taking the absolute value of the difference between the current motion acceleration and the standard acceleration. The obtained joint angle deviation Dθ, muscle activity deviation Deg, movement rhythm deviation Dv, and movement impact deviation Da are integrated to calculate the overall movement deviation degree MD. The method for obtaining the overall movement deviation MD is as follows: square the joint angle deviation Dθ, muscle activity deviation Deg, movement rhythm deviation Dv, and movement impact deviation Da respectively, sum them up, and then take the square root of the summed value to obtain the overall movement deviation MD. The feedback intensity value (Feed) is obtained by dividing the overall motion deviation (MD) by 1 and summing it with the absolute value of the motion evaluation error (E). Analyze the feedback intensity values ​​(Feed) and classify the feedback levels; The feedback levels are categorized as follows: When the feedback intensity value Feed < 0.2, it indicates the first feedback level, with no prompt. When 0.2 ≤ Feed intensity value ≤ 0.5, it indicates the second feedback level, and a prompt will appear on the screen. When 0.5 < Feed intensity value Feed ≤ 1, it indicates the third feedback level, which includes visual and vibration cues; When 1 < Feed intensity value Feed, it indicates the fourth feedback level, a strong vibration warning.

7. The orthopedic postoperative rehabilitation training monitoring system based on intelligent sensing according to claim 6, characterized in that: The feedback output control unit generates specific feedback signals based on the feedback intensity Feed, including the feedback signal amplitude sf, the feedback vibration intensity Vb, and the visual cue level Lv; The feedback signal amplitude sf is obtained as follows: Sf = ln(1 + Feed); where ln represents the logarithmic function; The feedback vibration intensity Vb is obtained as follows: Vb = Vmax × Sf; where Vmax represents the maximum vibration intensity of the equipment. The visual cue level Lv is obtained as follows: Lv = ⌊3 × Sf⌋; where ⌊⌋ represents the floor function. Get different screen cues based on the visual cue level (Lv): When the visual cue level Lv=0, there is no cue; When the visual cue level Lv=1, a yellow warning is displayed on the screen. When the visual cue level Lv=2, the screen displays an orange cue. When the visual cue level Lv=3, a red cue is displayed on the screen.

8. The orthopedic postoperative rehabilitation training monitoring system based on intelligent sensing according to claim 7, characterized in that: The automatic adjustment module for rehabilitation training plans includes a training status assessment unit and a training plan adjustment unit. The training status assessment unit calculates the patient's current training status index SQ based on the feedback intensity Feed and the movement assessment error E, and determines whether the patient's current training is in a stable state, a deviation state, or an insufficient load state. The training state index SQ is obtained by taking the square root of the sum of the square of the feedback intensity Feed and the square of the action evaluation error E. When the training state index SQ ≤ 0.7, it indicates that the training action is stable; When the training state index SQ > 0.7, it indicates that the action deviates and the training action is abnormal; The motion stability KQ is further calculated based on the training state index SQ. Action stability KQ is obtained by dividing 1 by 1 and summing it with the training state index SQ; The training plan adjustment unit automatically adjusts the training time and intensity based on the training state index SQ and the movement stability KQ, obtains the adjusted training time and intensity, and generates a new training plan. The adjusted training time is obtained by multiplying the current training time by the movement stability KQ; The adjusted training intensity is obtained by dividing the current training intensity by 1 and summing it with the training state index SQ; Based on the adjusted training time and intensity, calculate the training rhythm correction amount R and revise the training plan accordingly. The training pacing adjustment R is obtained by multiplying the current training time and the current training intensity by the product of the adjusted training time and the adjusted training intensity. When the training pace correction R < 0.5, it indicates that the training load is reduced and the rest time is increased. When 0.5 ≤ training pace correction R ≤ 1.2, it means that the training plan will remain in its current adjusted state; When 1.2 < training pace correction R, the training state is stable, and the training task is gradually increased in subsequent training cycles.

9. The orthopedic postoperative rehabilitation training monitoring system based on intelligent sensing according to claim 8, characterized in that: The data storage and analysis report module includes a data encapsulation and cloud storage unit and a rehabilitation progress analysis and report generation unit; The data encapsulation and cloud storage unit receives the generated training data, including the monitoring dataset KW, action evaluation error E, feedback intensity Feed, and training plan adjustment results; First, all types of data are sorted according to timestamps and a training time series is established. Then, the data is divided into several training segments based on continuous action changes, and the corresponding action error, feedback triggering situation, and training plan adjustment information are recorded in each segment. Next, a data index is created for each training segment, forming a training record structure that includes segment number, time range, action parameters, and feedback events; Finally, the packaged training records are uploaded to the cloud database and stored in layers according to the original data layer, training segment layer and event index layer. The rehabilitation progress analysis and report generation unit aggregates training segments based on training dates or training cycles, and calculates the training duration, number of movement segments, and number of feedback triggers within each cycle. Subsequently, the joint angle data in each segment were statistically processed to obtain the maximum and minimum angles of each segment, and the range of motion of the joints was obtained accordingly; at the same time, segment-level statistics were performed on the motion evaluation error sequence to count the duration and frequency of errors. Then, by combining the feedback intensity and training plan adjustment records, the deviations that occurred during the training process and the corresponding adjustment behaviors are analyzed; finally, a rehabilitation progress report is generated based on changes in range of motion, error changes, and training adjustments. The report includes range of motion, training stability, recovery progress, and subsequent training recommendations.

10. A postoperative orthopedic rehabilitation training monitoring device based on intelligent sensing, applied to the postoperative orthopedic rehabilitation training monitoring system based on intelligent sensing as described in any one of claims 1 to 9, characterized in that: It includes an inertial sensor module, a surface electromyography (EMG) acquisition module, a flexible angle sensor, a vibration feedback module, and a mobile data terminal.

Images