Intelligent assessment method and system for postoperative rehabilitation of traumatic orthopedics department

Abstract

The invention relates to the technical field of rehabilitation assessment, in particular to an intelligent assessment method and system for postoperative rehabilitation of traumatic orthopedics, and the method comprises the following steps: calling an Internet of Things and a myoelectricity sensor to collect key motion parameters, analyzing abnormal fluctuation and response offset, recognizing muscle group collaborative abnormality, quantifying trajectory deviation and standardizing scores, and carrying out intelligent assessment on postoperative rehabilitation of traumatic orthopedics. And combining training and compliance sequence comparison fluctuation, adjusting a rehabilitation plan, and outputting a behavior management record. According to the invention, by fusing the joint posture and the myoelectricity activation data, constructing a symmetric matching mechanism between action execution and muscle group response, and combining action mutation detection and offset trend analysis, accurate recognition and quantitative judgment of an abnormal action structure are realized, the recognizability of abnormal behaviors in the rehabilitation process is improved, and the rehabilitation effect is improved. The response sensitivity of the evaluation index to the training compliance fluctuation is enhanced, and the adaptation ability of the rehabilitation plan in the scenes of asymmetric actions and discontinuous behaviors is optimized, so that the evaluation result fits the real rehabilitation state of the patient.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of rehabilitation assessment, in particular to a trauma orthopedic postoperative rehabilitation intelligent assessment method and system. BACKGROUND

[0002] The technical field of rehabilitation assessment includes related methods and systems for measuring and judging the physical function state of patients during postoperative or disease recovery period. By monitoring human motion function, physiological parameters, neural response indicators, combining medical standards and evaluation scales, the rehabilitation process is quantitatively analyzed and the state is recognized. Through wearable sensors, biological parameter detection equipment and special evaluation software, the patient's action amplitude, gait change, joint activity, muscle strength and other aspects are comprehensively evaluated to provide rehabilitation monitoring data support for medical personnel. Combined with multi-dimensional physiological and motion data acquisition mechanism, the corresponding quantitative index model is established, and the standardized evaluation process is formed to support clinical rehabilitation management.

[0003] Among them, the trauma orthopedic postoperative rehabilitation intelligent assessment method refers to setting functional indicators related to trauma orthopedic during the patient's rehabilitation process, using multiple sensors to monitor bone joint activity, gait posture change, muscle electrical signal response and other elements in real time, and transmitting the collected data to the operation unit for feature extraction and index quantitative analysis. Relying on the inertial measurement unit, the pressure sensor and the surface electromyography acquisition device, the dynamic data of the target regions such as the knee joint, ankle joint and hip joint are distributedly sampled, and then according to the set motion task standard and rehabilitation scale, the data is mapped into the rehabilitation scoring model to complete the grading evaluation. In the evaluation process, the collected results are processed and output evaluation indexes according to the fixed rules such as action amplitude recognition algorithm, motion posture matching process and electromyography response window extraction program, which are used to judge the functional recovery state of the patient.

[0004] Although the traditional orthopedic postoperative rehabilitation intelligent assessment technology has the basic measurement capability of joint range of motion, muscle strength and posture change, it cannot deeply identify the time offset structure of the motion behavior, lacks the dynamic analysis capability of bilateral muscle group coordination and action symmetry relationship, and cannot form the response index strongly related to the rehabilitation state change under the conditions of insufficient training task continuity and unstable action execution. It is easy to cause state evaluation delay or abnormal recognition omission, which is manifested in actual application as abnormal action can only be recognized when the amplitude deviation is obvious, the recognition coverage of behavior rhythm disorder or training compliance fluctuation is insufficient, which limits the response efficiency of evaluation feedback in dynamic management scenarios and reduces the applicability and timeliness of individualized training adjustment strategies. SUMMARY

[0005] The purpose of the present application is to solve the shortcomings in the prior art and to provide a trauma orthopedic postoperative rehabilitation intelligent assessment method and system.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: an intelligent assessment method for postoperative rehabilitation in trauma orthopedic surgery, comprising the following steps: S1: Call IoT sensors to collect the angular velocity sequence of the knee and ankle joints during the squatting action, calculate the angular velocity difference sequence in the main rotation direction, compare it with the angular velocity change range of the same node in the previous cycle, analyze the action mutation rate, detect abnormal monitoring data, and output the abnormal data identification result; S2: Call the abnormal data identification results, combine them with the signals collected by the electromyography sensor, compare the upward trend of the electromyography signal in the initiation phase of the movement with the acceleration direction of the corresponding joint angle change, determine the degree of coordination between the muscle group activation behavior and the joint movement output response, and generate response offset information. S3: Using the response offset information, compare the activation timing and amplitude ratio of the left and right quadriceps muscles in the alternating leg raise training task, combine the change direction of the X-axis posture angle, analyze the symmetrical coordination structure between muscle group activation and posture response, identify structural abnormal movements, and generate muscle group synergistic analysis results. S4: Based on the muscle group synergy analysis results, screen the multi-cycle point set of the joint three-dimensional trajectory in continuous hip abduction training, calculate the deviation from the standard path, determine the frequency of abnormal nodes, adjust the weight of the movement standardization score, and output the postoperative rehabilitation score.

[0007] As a further aspect of the present invention, the abnormal data identification results include angular velocity fluctuation amplitude, continuous time mutation points, and abnormal segments of the monitoring sequence; the response offset information specifically includes the direction of change of electromyographic upward trend, angular acceleration offset segments, and the proportion of response mismatch periods; the muscle group synergy analysis results include activation time difference, activation intensity ratio, and posture coordination direction; and the postoperative rehabilitation score specifically refers to the trajectory repetition offset frequency, score weight adjustment coefficient, and movement standardization score interval.

[0008] As a further aspect of the present invention, the step of obtaining the abnormal data identification result specifically includes: S111: Call IoT sensors to collect angular velocity data of the knee and ankle joints during continuous squatting movements, extract the angular velocity signal sequence of the main rotation direction in each cycle, arrange the angular velocity point sets in multiple cycles in chronological order, and generate a continuous angular velocity sequence group. S112: Based on the continuous angular velocity sequence group, the angular velocity mutation rate index is calculated by comparing the angular velocity change range of the same node in the previous period, and the angular velocity mutation distribution is obtained by statistically analyzing the fluctuation amplitude of each node. S113: Call the angular velocity mutation distribution, determine the abnormal change trend of the fluctuation amplitude in the time continuous structure, analyze the action mutation ratio, detect abnormal monitoring data, and output the abnormal data identification result.

[0009] As a further aspect of the present invention, the step of obtaining the response offset information specifically includes: S211: Call the abnormal data identification results, collect the signal sequence of the electromyography sensor during the patient's movement initiation phase, fit the slope of the extracted signal rising segment, calculate the electromyography activation trend change rate in each cycle, and extract the acceleration vector by combining the synchronously collected joint angle change information to obtain the electromyography trend response structure. S212: Based on the electromyographic trend response structure, compare the degree of coordination between the activation rate change of the muscle group corresponding to each joint within the same period and the acceleration direction of the target joint, determine the transient deviation structure within the period, calculate the response coordination matching index, and obtain the trend matching fluctuation segment index sequence. S213: Call the trend matching fluctuation segment index sequence to determine the degree of response coordination between muscle group activation behavior and target joint movement output, and identify the proportion of the output response mismatch segment between muscle group nerve activation and movement execution in the time series, and generate response offset information.

[0010] As a further aspect of the present invention, the steps for obtaining the muscle group synergy analysis results are specifically as follows: S311: Based on the response offset information, extract the activation frame indexes of the left and right quadriceps muscles within the alternating leg raise training cycle, compare the time intervals of the activation frames on the left and right sides, determine the degree of activation timing offset, and obtain the activation timing offset index. S312: Based on the activation timing offset index, extract the normalized peak values ​​of electromyography on the left and right sides in each cycle, and calculate the muscle group posture symmetry coordination index by combining the X-axis posture angle, muscle group activation duration, and activation frame position difference of the corresponding cycle. S313: Call the muscle group posture symmetry coordination index, combine it with the activation time sequence offset index, determine the symmetry state of the action execution, determine the periodic structural abnormal action, and obtain the muscle group coordination analysis results.

[0011] As a further aspect of the present invention, the step of obtaining the postoperative rehabilitation score specifically includes: S411: Based on the muscle group synergy analysis results, screen the three-dimensional trajectory point set of the joint in each cycle of the continuous hip abduction training task, extract the key trajectory nodes in each cycle, and obtain the set of key trajectory nodes. S412: Call the set of key nodes of the trajectory, compare the distance between the key nodes of each cycle trajectory and the corresponding points of the standard action path, determine the frequency of occurrence of abnormal deviation nodes in each cycle, and obtain the frequency index of abnormal deviation nodes. S413: Based on the abnormal deviation node frequency index, analyze the repetitive deviation trend of the spatial trajectory in periodic training, and adjust the corresponding weight of the action standardization in the scoring structure in combination with the frequency of abnormal node occurrence, and output the postoperative rehabilitation score.

[0012] As a further aspect of the present invention, the method further includes: S5: Using the postoperative rehabilitation score, analyze the patient's daily training task completion record, action execution frame coverage, and task interval time distribution, construct a compliance index sequence, compare it with the rehabilitation score sequence, analyze the impact of patient behavior fluctuations on functional rehabilitation, adjust the patient's rehabilitation plan, and output the patient behavior management record. The patient behavior management records include the percentage of training task completion, the percentage of action frame coverage, and the frequency of training interruptions.

[0013] As a further aspect of the present invention, the steps for obtaining the patient behavior management record are specifically as follows: S511: Using the postoperative rehabilitation score, analyze the patient's daily training task completion record, action execution frame coverage data and training task interval distribution to obtain the daily actual number of completions, action segment coverage rate and training interruption days, and obtain a set of task completion statistical indicators. S512: Based on the task completion statistical indicator set, calculate the percentage of actual completion times, the coverage rate of continuous action segments, and the frequency of training interruption days, construct a compliance indicator sequence, and compare it with the rehabilitation score sequence to obtain compliance fluctuation correlation indicators. S513: Based on the aforementioned compliance fluctuation correlation index, identify the impact of compliance changes on rehabilitation scores, adjust the patient's rehabilitation plan, and output patient behavior management records.

[0014] A smart assessment system for postoperative rehabilitation of trauma orthopedic patients, the system being used to execute the aforementioned smart assessment method for postoperative rehabilitation of trauma orthopedic patients, the system comprising: The abnormal data identification module calls IoT sensors to collect the angular velocity sequence of the knee and ankle joints during the squatting action, calculates the angular velocity difference sequence in the main rotation direction, compares it with the angular velocity change range of the same node in the previous cycle, analyzes the action mutation rate, detects abnormal monitoring data, and outputs the abnormal data identification results. The muscle group response coordination module calls the abnormal data identification results, combines them with the signals collected by the electromyography sensor, compares the upward trend of the electromyography signal in the initiation phase of the movement with the acceleration direction of the corresponding joint angle change, judges the degree of coordination between the muscle group activation behavior and the joint movement output, and generates response offset information. The muscle group synergy analysis module uses the response offset information to compare the activation timing and amplitude ratio of the left and right quadriceps muscles in the alternating leg raise training task. Combined with the change direction of the X-axis posture angle, it analyzes the symmetrical coordination structure between muscle group activation and posture response, identifies structural abnormal movements, and generates muscle group synergy analysis results. Based on the muscle group synergy analysis results, the motion trajectory assessment module filters the multi-cycle point set of the joint three-dimensional trajectory in continuous hip abduction training, calculates the deviation from the standard path, judges the frequency of abnormal nodes, adjusts the weight of the movement standardization score, and outputs the postoperative rehabilitation score. The patient behavior management module uses the postoperative rehabilitation score to analyze the patient's daily training task completion records, action execution frame coverage, and task interval time distribution, constructs a compliance index sequence, compares it with the rehabilitation score sequence, analyzes the impact of patient behavior fluctuations on functional rehabilitation, adjusts the patient's rehabilitation plan, and outputs patient behavior management records.

[0015] Compared with the prior art, the advantages and positive effects of the present invention are as follows: In this invention, by fusing joint posture and electromyographic activation data, a symmetrical matching mechanism between movement execution and muscle group response is constructed. Combined with movement mutation detection and deviation trend analysis, the accurate identification and quantitative judgment of abnormal movement structures are achieved, which improves the identifiability of abnormal behaviors during rehabilitation, enhances the sensitivity of assessment indicators to fluctuations in training compliance, optimizes the adaptability of rehabilitation plans in scenarios of movement asymmetry and behavioral discontinuity, and makes the assessment results more consistent with the patient's actual rehabilitation status. Attached Figure Description

[0016] Figure 1 This is a schematic diagram of the main steps of the present invention; Figure 2 This is a flowchart of the abnormal data identification result acquisition process of the present invention; Figure 3 This is a flowchart of the response offset information acquisition process of the present invention; Figure 4 This is a flowchart of the process for obtaining the muscle group synergy analysis results of the present invention; Figure 5 This is a flowchart of the postoperative rehabilitation score acquisition process of the present invention; Figure 6 This is a flowchart of the patient behavior management record acquisition process of the present invention. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0018] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0019] Please see Figure 1 This invention provides a technical solution: an intelligent assessment method for postoperative rehabilitation in trauma orthopedic surgery, comprising the following steps: S1: Call IoT sensors to collect the angular velocity sequence of the knee and ankle joints during the squatting action, calculate the angular velocity difference sequence in the main rotation direction, compare it with the angular velocity change range of the same node in the previous cycle, analyze the action mutation rate, detect abnormal monitoring data, and output the abnormal data identification result; S2: Call the abnormal data identification results, combine them with the signals collected by the electromyography sensor, compare the upward trend of the electromyography signal in the initiation phase of the movement with the acceleration direction of the corresponding joint angle change, determine the degree of coordination between the muscle group activation behavior and the joint movement output response, and generate response offset information. S3: Using response offset information, compare the activation timing and amplitude ratio of the left and right quadriceps muscles in the alternating leg raise training task. Combined with the direction of change of X-axis posture angle, analyze the symmetrical coordination structure between muscle group activation and posture response, identify structural abnormal movements, and generate muscle group synergy analysis results. S4: Based on the results of muscle group synergy analysis, screen the multi-cycle point set of the joint three-dimensional trajectory in continuous hip abduction training, calculate the deviation from the standard path, determine the frequency of abnormal nodes, adjust the weight of the movement standardization score, and output the postoperative rehabilitation score. S5: Using postoperative rehabilitation scores, analyze the patient's daily training task completion records, action execution frame coverage, and task interval time distribution to construct a compliance index sequence. Compare this sequence with the rehabilitation score sequence to analyze the impact of patient behavior fluctuations on functional rehabilitation, adjust the patient's rehabilitation plan, and output the patient behavior management record.

[0020] Abnormal data identification results include angular velocity fluctuation amplitude, continuous time mutation points, and abnormal segments in the monitoring sequence. Response offset information specifically includes the direction of change in the upward trend of electromyography, angular acceleration offset segments, and the proportion of response mismatch periods. Muscle group synergy analysis results include activation time difference, activation intensity ratio, and posture coordination direction. Postoperative rehabilitation scores specifically refer to the frequency of trajectory repetition offset, score weight adjustment coefficient, and movement standardization score interval. Patient behavior management records include the percentage of training task completion, movement frame coverage ratio, and training interruption frequency.

[0021] Please see Figure 2 The specific steps for obtaining the abnormal data identification results are as follows: S111: Call IoT sensors to collect angular velocity data of the knee and ankle joints during continuous squatting movements, extract the angular velocity signal sequence of the main rotation direction in each cycle, arrange the angular velocity point sets in multiple cycles in chronological order, and generate a continuous angular velocity sequence group. As the patient performs a squatting motion, each cycle is divided into a fixed number of sampling points according to the time sequence. In this embodiment, each cycle is divided into 20 equally spaced nodes. The sensor collects the corresponding raw angular velocity value at each node. For example, in the first cycle, the angular velocities of the knee joint node are as follows: (Unit: ° / s), the angular velocity of the ankle joint node is (Unit: ° / s) This set of raw data is arranged according to the node order to form the angular velocity sequence for this period. Then, the main rotation direction is extracted. Since the knee joint is the main load-bearing joint during squatting, the knee joint is selected as the main rotation direction angular velocity signal. In each period, the angular velocity values ​​of the 20 nodes are subjected to maximum and minimum normalization. Assuming that the maximum value of the knee joint angular velocity in this period is 75° / s and the minimum value is -2° / s, the normalized value of the first node is calculated as follows: Similarly, the normalized values ​​of other nodes are calculated to obtain a complete normalized sequence. Then, the normalized results from multiple consecutive periods are spliced ​​together to generate a continuous sequence of angular velocities. In the second and third periods, the normalized sequences are obtained in the same way and combined into a three-dimensional array. In the array index, u represents the joint number. In this example, u=1 represents the knee joint, u=2 represents the ankle joint, v represents the time node number from 1 to 20, and t represents the current cycle number. This process ensures that the node data of each cycle can be arranged into the time continuous structure, and finally obtains the angular velocity continuous sequence group, which serves as the basis for subsequent calculations and comparisons.

[0022] S112: Based on a continuous sequence of angular velocities, the formula is used to compare the range of angular velocity changes at the same nodes in the previous period with the range of angular velocity changes at the same nodes. ; Calculate the angular velocity mutation rate index and obtain the angular velocity mutation distribution by statistically analyzing the fluctuation amplitude of each node; in, For the current cycle of joints At the time point The normalized angular velocity value is obtained by collecting the raw data of the angular velocity in the main rotation direction during the squatting motion and performing a maximum-minimum normalization transformation on each node in each cycle. For the previous cycle, the lower joint At the time point The normalized value of the angular velocity is obtained by normalizing the raw angular velocity data of the same node in the previous cycle. For the next cycle of the lower joint At the time point The normalized value of the angular velocity is obtained by normalizing the raw angular velocity data of the same node in the previous cycle. To collect joint numbers, indicating the knee or ankle joint, This serves as the index for time nodes within the current period, dividing the time series into a fixed number of equally spaced nodes within each period. , , These represent the action execution cycle numbers for the current cycle, the previous cycle, and the cycle before that, respectively. The cycle is segmented based on the number of consecutive frames captured during action acquisition. For joints At the time point The angular velocity mutation rate index represents the degree of angular velocity fluctuation of the corresponding node of the same joint within a continuous period; Based on the comparison of the period based on the continuous sequence of angular velocities, the normalized values ​​of the angular velocities at the same node in the current period, the previous period, and the period before that are first extracted and denoted as follows: , , For example, at the 5th node, assuming the normalized values ​​for the knee joint in the three cycles are 0.85, 0.80, and 0.78 respectively, substituting them into the calculation formula: ; The numerator is the sum of the absolute values ​​of the period differences, i.e.: ; The denominator is the square root of the sum of the three terms, i.e.: , result: ; This value is the angular velocity mutation rate index of the knee joint at node 5. The same calculation is performed on all nodes sequentially to form a complete mutation rate distribution. The angular velocity mutation rate index is a quantitative expression of the trend of angular velocity change of a specific joint at the same moment in a multi-cycle continuous squatting movement over three consecutive cycles. The larger the value, the more concentrated or sudden the drastic fluctuation of angular velocity at that node. This index is used to accurately locate abnormal points in the movement process. By correlating with spatial location data, it can identify non-physiological movement deviation segments suspected to be caused by postural deviations, sudden pauses, or joint vibrations. As a core calculation item for triggering the abnormality monitoring mechanism, the index can be used to drive subsequent data acquisition and calibration, training task reset, or feedback to users to adjust their movements, thereby ensuring the continuity and standardization of data acquisition in rehabilitation training, stabilizing the movement judgment process, improving assessment accuracy, and reducing the false trigger rate. In the formula, the superscript 't' represents the period number, the subscript 'u' represents the joint number, the subscript 'v' represents the node number, the vertical bar indicates absolute value operation, and the square root symbol indicates square root operation. The calculation logic is to represent the fluctuation amplitude by the absolute value of the difference between consecutive periods, and then normalize and scale it through the denominator. For ease of explanation, the data results for some nodes are listed below:

[0023] As shown in Table 1, node 15 has the highest mutation rate, which is 0.064, indicating that the fluctuation range of the knee joint angular velocity is the largest at this node.

[0024] In practical applications, to determine whether the mutation rate is abnormal, a baseline interval needs to be set. For example, through statistical analysis of 20 recovered patients, the mutation rate distribution within the normal range of the knee joint was obtained. A value exceeding 0.05 is considered an abnormal fluctuation. In this example, the result for node 15, 0.064, is higher than this threshold and is therefore marked as an abnormal node. By introducing the normalized difference of nodes over three consecutive cycles, the formula can highlight the degree of abrupt change in the movement rhythm at the numerical level. Combined with the adjustment of the normalized denominator, it avoids interference from excessively large differences in the absolute value of angular velocity among different patients or in different cycles on the mutation rate index.

[0025] S113: Call the angular velocity mutation distribution to determine the abnormal change trend of the fluctuation amplitude in the time continuous structure, analyze the action mutation ratio, detect abnormal monitoring data, and output the abnormal data identification results. After obtaining the mutation distribution, it is necessary to further determine the fluctuation amplitude in the continuous time structure. The specific process involves sequentially retrieving the mutation rate index of each node and comparing it with the previously set threshold range. When a node's value exceeds 0.05, it is marked as an anomalous node, and the number and location of anomalous nodes are counted. For example, in the results of Table 1, node 15 is an anomalous point, so the number of anomalous nodes in this period is 1. This count is then repeated in the next period. If anomalies occur at adjacent nodes in two consecutive periods, a mutation trend is identified. Assuming that node 16 in the second period also shows an index value of 0.067, then nodes 15 and 16 are both anomalous in two consecutive periods, and this is recorded as a mutation trend range. Then, we calculate the proportion of this trend interval to the total number of nodes in the entire cycle. The mutation rate is used as the action mutation ratio. If the mutation rate exceeds 0.08, it is considered that there is a significant abnormal action in that period. In this example, the ratio is 0.1, which exceeds the threshold. Therefore, the abnormal monitoring data is output, and the abnormal segment index and mutation rate value are recorded in the results. The final abnormal data identification results are: the mutation rate of the first period is 0.1, and the abnormal nodes are concentrated in nodes 15 and 16.

[0026] Please see Figure 3 The specific steps for obtaining the response offset information are as follows: S211: Call the abnormal data identification results, collect the signal sequence of the electromyography sensor during the patient's movement initiation phase, fit the slope of the extracted signal rising segment, calculate the rate of change of electromyography activation trend in each cycle, and extract the acceleration vector by combining the synchronously collected joint angle change information to obtain the electromyography trend response structure. The abnormal data identification results were retrieved, specifically a mutation rate of 0.1 in cycle 1 with abnormalities at nodes 15 and 16. Based on this, electromyographic (EMG) signal sequences were collected during the patient's squatting initiation phase, defined as the first 10 frames of the movement. The signal amplitudes obtained using an EMG sensor at the quadriceps muscle were as follows: (Unit: mV), before normalization, the maximum value of this data set is 1.00 and the minimum value is 0.12. After normalization, we get... In this normalized sequence, the rising segment is selected for slope fitting. Specifically, the calculation involves performing a difference operation on each pair of adjacent frames and dividing by the inter-frame interval. In this example, the inter-frame interval is 0.02 seconds. The slope of the first two frames... Then the slope of the second and third frames Based on this, the slope of each subsequent segment is calculated to obtain the trend change rate sequence. Then, combining the joint angle change information, the knee joint angle change is collected in the same steps, from the initial 90° to squatting to 60°, with a change of -3° per frame. The acceleration vector direction value is obtained by difference, and the direction sequence is obtained after normalization. The normalized electromyographic trend and acceleration vector were aligned in the time frame sequence. Specifically, the normalized electromyographic value was 0.38 in the 5th frame, corresponding to an acceleration direction of -0.80, and the normalized electromyographic value was 0.81 in the 8th frame, corresponding to an acceleration direction of -0.65. The electromyographic trend response structure was finally obtained, providing basic data for subsequent matching degree calculation.

[0027] S212: Based on the electromyographic trend response structure, compare the degree of coordination between the activation rate change of the corresponding muscle group at each joint and the acceleration direction of the target joint within the same cycle to determine the transient deviation structure within the cycle, using the formula: ; Calculate the response coordination matching degree index to obtain the trend matching fluctuation segment index sequence; in, For the first In the period, the 1st The response coordination matching index of each joint is obtained by calculating the degree of coupling between the electromyographic activation amplitude and the acceleration direction. For the first In the period, the 1st The muscle groups corresponding to the joints in the first... The normalized electromyographic activation amplitude at each frame time is obtained by acquiring the rising segment of the electromyographic signal and performing a normalization operation. For the first The joint in the first The normalized acceleration direction value at each frame time is obtained by collecting angle change data and combining it with the velocity change to calculate the normalized direction vector. The total number of frames within a period is obtained by the difference between the start and end frame indices of the period. The frame sequence number index indicates the position of the current frame within the period. The period number index indicates the periodic position of the action in the time series. The joint number index indicates the target joint controlled by the muscle group; Based on the established electromyographic trend response structure, it is necessary to compare the degree of coordination between the activation rate change of the corresponding muscle group at each joint and the acceleration direction of the target joint within the same period. The calculation formula is as follows: ; in, This represents the response coordination degree of the j-th joint in the i-th cycle. The numerator is the sum of the products of the electromyographic activation amplitude and the direction of joint acceleration in each frame, taking the absolute value. The denominator is the square root of the sum of the squares of the electromyographic activation amplitude and the squares of the acceleration direction, plus 1 to prevent the denominator from approaching 0. Specifically, assuming M=10 frames for the knee joint in the first cycle, the normalized electromyographic value is... for Joint acceleration direction for The sequence of products is obtained by multiplying frame by frame. Adding each term together, we get a sum of -3.1785. Taking the absolute value of 3.1785 as the numerator, we then calculate the denominator, frame by frame. , such as the first frame The second frame The summation frame by frame is 8.1769, the square root is 2.86, and adding 1 gives 3.86. Therefore, the formula result is: ; This value represents the knee joint's response coordination matching degree in the first cycle. The response coordination matching degree index quantifies the temporal and directional consistency between the electromyographic activation behavior of the target muscle group and the motor output generated by the joint it controls within a specific cycle. Specifically, it reflects whether the activation of neural signals can effectively drive the joint to output in the expected acceleration direction. A high index value indicates a good neuromuscular functional transmission relationship between the activation behavior of the muscle group and the joint response within that cycle; a low index value may indicate activation lag, output delay, or inverse vector mismatch, suggesting a risk of mismatch within that cycle. This index, as a core parameter for quantifying output response coordination, can be used as a refined input basis for behavioral fluctuation analysis, functional impairment assessment, and subsequent intervention decisions.

[0028]

[0029] As shown in Table 2, most of the multiplication results of the frame sequences are negative, reflecting a certain degree of inverse relationship between the electromyographic (EMG) signal and the acceleration direction. However, the coordination matching index obtained after taking the absolute value is 0.823. Combined with statistical experiments, the threshold range of 0.75 to 0.90 is determined to be a normal match. This example falls within this range, thus generating a normal record in the trend matching fluctuation segment index sequence. The formula couples the time-dimensional EMG and acceleration data into a single value by multiplying frame by frame, summing the results, and then taking the absolute value. This avoids the interference of random fluctuations in a single frame on the overall matching degree, thereby characterizing the degree of response coordination at the overall level.

[0030] S213: Call the trend matching fluctuation segment index sequence to determine the degree of response coordination between muscle group activation behavior and target joint motion output, and identify the proportion of output response mismatch segments between muscle group nerve activation and motion execution in the time series, generating response offset information; After obtaining the trend-matching fluctuation segment indicator sequence, further analysis of the matching degree within the time series is required. Specifically, this involves period-by-period and joint-by-joint retrieval. The value is when the result is within the normal range. The time stamp is marked as a normal match. When it is below 0.75, it is judged as insufficient deviation. When it is above 0.90, it is judged as excessive deviation. In this embodiment, the matching degree of the knee joint in the first cycle is 0.823, which falls within the normal range. Therefore, it is not marked as mismatch. However, if it is 0.72 when calculated in the second cycle, it is judged as an insufficient deviation interval. The mismatch segment index of this cycle is recorded and the proportion is calculated. Assuming that this situation occurs in 3 frames out of the 10 frames, the proportion is 0.3. Then, the cycle number = 2 and the mismatch proportion = 0.3 are recorded in the response offset information. This data is summarized into the output result. The final generated response offset information is: normal in the first cycle, mismatch proportion of 0.3 in the second cycle. Subsequent data continue to be accumulated and recorded according to this rule.

[0031] Please see Figure 4 The specific steps for obtaining the results of the muscle group synergy analysis are as follows: S311: Based on the response offset information, extract the activation frame indexes of the left and right quadriceps muscles within the alternating leg raise training cycle, compare the time intervals of the activation frames on the left and right sides, determine the degree of activation timing offset, and obtain the activation timing offset index. Using response offset information (i.e., normal in cycle 1, mismatch rate of 0.3% in cycle 2), activation frame indices of the left and right quadriceps muscles in the alternating leg raise training task were extracted. Each complete movement cycle of this task was set to contain 20 frames. The activation frame of the left quadriceps in cycle 1 was frame 6, and the activation frame of the right quadriceps in cycle 1 was frame 8, with a time interval of 2 frames. If the interval between each frame was 0.02s, the corresponding time difference was 0.04s, resulting in an activation timing offset of 0.04s for cycle 1. In cycle 2, the activation frame of the left quadriceps was frame 5, and the activation frame of the right quadriceps was frame 10, with a time interval of 5 frames and a time difference of 0.10s, resulting in an activation timing offset of 0.10s for cycle 2. The offset values ​​for all cycles were recorded sequentially and compared with a threshold range. The threshold was set with reference to the range of quadriceps activation time delay in clinical rehabilitation literature, with the normal range being [missing information]. A delay of less than 0.02s is defined as no significant delay, and a delay of more than 0.06s is defined as abnormal delay. In this example, the offset of the first cycle is 0.04s, which is within the normal range. The offset of the second cycle is 0.10s, which exceeds the threshold and is marked as an abnormal cycle. The final activation timing offset index is: the first cycle is normal and the second cycle is abnormal.

[0032] S312: Based on the activation time offset index, extract the normalized peak values ​​of left and right electromyography in each cycle. Combined with the X-axis pose angle, muscle group activation duration, and activation frame position difference of the corresponding cycle, the formula is used: ; Calculate the muscle group posture symmetry coordination index; in, For the first The normalized electromyographic peak value of the left quadriceps femoris muscle in each cycle was obtained by collecting the electromyographic signal of the left side, extracting the maximum activation potential, and then normalizing it. For the first The normalized electromyographic peak value of the right quadriceps femoris muscle in each cycle was obtained by acquiring the right electromyographic signal, extracting the maximum activation potential, and then normalizing it. For the first The normalized value of the X-axis attitude angle within each cycle is obtained by collecting the X-axis attitude angle output by the inertial measurement unit and normalizing it. For the first The normalized value of the duration of muscle group activation within each cycle is obtained by calculating and normalizing the difference between the start and end frame positions of electromyographic activation. For the first The normalized value of the difference between the positions of the left and right active frames within a period is obtained by comparing the index positions of the left and right active frames and normalizing them. The total number of training cycles is obtained by counting the number of complete movement cycles in the alternating leg raise exercise. The cycle index indicates the sequence number of the action cycle currently being calculated. It is a muscle group posture symmetry coordination index, used to measure the degree of coordination between muscle group activation response and posture structure; Using the activation time offset index, the normalized peak electromyography (EMG) values ​​of the left and right quadriceps muscles in each cycle were extracted. Combined with the X-axis pose angle, muscle activation duration, and activation frame position difference of the corresponding cycle, the muscle group pose symmetry coordination index was calculated using a formula. ; In the formula, The percentage difference between the left and right peak values ​​is used to characterize the degree of imbalance between the two peak values. The normalized values ​​for the attitude angles are: 5° for the first cycle (normalized to 0.50), 0.80 for the muscle group activation duration, and 0.10 for the difference between the left and right activation frame positions (2 frames in the first cycle, normalized to 0.10). These values ​​are then substituted into the formula for calculation, assuming the peak value is on the left side in the first cycle. Right peak The proportion of peak difference is The attitude angle term is 0.50, and the denominator term is... Substituting the values, we get the first period exponent term as follows: In the second cycle, the left peak Right peak The difference ratio is 0.136 / 1.62 = 0.084, the attitude angle is normalized to 0.55, the duration is normalized to 0.75, the frame difference is normalized to 0.25, and the denominator is... The exponent term is If the total period m=2, then the overall index is: ; The muscle group postural symmetry coordination index is an indicator used to quantify the coordination between the left and right muscle groups in terms of activation timing, output intensity, and postural control during alternating movement tasks. This dimensionless index reflects whether the quadriceps femoris forms a stable and symmetrical coordination structure across multiple training cycles. A smaller index value indicates that the activation intensity of the left and right muscle groups tends to be consistent within the same cycle, with smaller differences in activation time and more symmetrical postural control; conversely, a larger index value indicates instability, imbalance in force output, or postural deviation during movement, suggesting potential motor control disorders or structural execution abnormalities. Therefore, this index can not only be used to identify movement abnormalities but also provide quantitative feedback support for coordination training during rehabilitation, making it a key parameter for constructing muscle group synergy analysis.

[0033]

[0034] As shown in Table 3, averaging the results from the two periods yields an overall index S = 0.0465. Experimental verification confirms that the normal range is... The value of 0.0465 falls on the edge of the normal upper limit, indicating that it is close to the critical point in maintaining symmetry. The formula combines the peak ratio with the attitude angle factor, and then normalizes it with the duration and frame difference, so that the index can comprehensively reflect the overall coordination of bilateral muscle groups in amplitude, time and attitude, rather than being limited to a single parameter.

[0035] S313: Call the muscle group posture symmetry coordination index, combine it with the activation time sequence offset index, determine the symmetry state of the action execution, determine the periodic structural abnormal actions, and obtain the muscle group coordination analysis results. After obtaining the muscle group posture symmetry coordination index, a comprehensive judgment is made in conjunction with the activation timing offset index. During the execution process, the results of the first cycle are retrieved first. The activation timing offset index is normal, and the posture coordination index is 0.0345, which is within the normal range. Therefore, the symmetry of the first cycle is judged to be good. Then, the results of the second cycle are retrieved. The activation timing offset is abnormal, and the index is 0.0584, which is higher than the upper limit of the normal range. Therefore, it is judged as a structural abnormal movement. This cycle is marked as abnormal. The number and proportion of abnormal cycles are counted in the total results. In this embodiment, there is 1 abnormal cycle out of 2 cycles, accounting for 0.5%. This is output as the muscle group coordination analysis result. The final result is: the first cycle is normal, the second cycle is abnormal, and the overall abnormality rate is 0.5%.

[0036] Please see Figure 5 The specific steps for obtaining the postoperative rehabilitation score are as follows: S411: Based on the results of muscle group synergy analysis, screen the three-dimensional trajectory point set of the joint in each cycle of the continuous hip abduction training task, extract the key trajectory nodes in each cycle, and obtain the set of key trajectory nodes. Based on the input of the muscle group synergy analysis results, specifically, one abnormal cycle exists within two cycles of alternating leg raise training, with an abnormality rate of 0.5%. Building upon this result, trajectory data for each cycle of the patient's continuous hip abduction training task were screened. An inertial measurement unit (IMU) and a 3D motion capture device were used to jointly acquire the trajectory point set of the hip joint in 3D space. The number of frames acquired per cycle was set to 30, with each frame corresponding to a 3D coordinate point (x, y, z). For example, some trajectory points acquired in the first cycle were (25.0, 18.0, 90.0), (26.5, 18.7, 89.5), (28.0, 19.2, 88.9), and (29.5, 19.8, 88.3). After obtaining the complete point set, key nodes need to be extracted. The selection of key nodes... The principle is that the trajectory is divided into nodes at spatial turning points, extreme points, or within a period. In this embodiment, six key nodes are selected in each period, corresponding to the starting point, intermediate peak point, lowest point, and transition point, respectively. The key nodes selected in the first period are (25.0,18.0,90.0), (28.0,19.2,88.9), (30.5,20.0,88.0), (28.5,19.5,89.0), (26.0,18.5,89.5), and (25.0,18.0,90.0), which are recorded as the trajectory key node set. The same extraction operation is performed in the second period to obtain the trajectory key node set for the second period. Finally, the key node sets of all periods are summarized to provide basic data for subsequent comparative analysis.

[0037] S412: Call the set of key nodes of the trajectory, compare the distance between the key nodes of each cycle trajectory and the corresponding points of the standard action path, determine the frequency of occurrence of abnormal deviation nodes in each cycle, and obtain the frequency index of abnormal deviation nodes. After retrieving the set of key nodes for the trajectory, it is necessary to compare the distances point-by-point between the key nodes of each cycle trajectory and the corresponding nodes of the standard movement path. The standard movement path is provided by a professional rehabilitation training database, and its coordinates at the same node are set as (25.0, 18.0, 90.0), (27.5, 19.0, 89.0), (30.0, 20.0, 88.0), (28.0, 19.5, 89.0), (26.0, 18.5, 89.5), and (25.0, 18.0, 90.0). The distance between the actual point and the standard point is calculated using the Euclidean distance formula. In the first cycle, at node 2, the actual coordinates are (28.0, 19.2, 88.9), and the standard coordinates are (27.5, 19.0, 89.0). The difference is (0.5, 0.2, -0.1), and the sum of squares is 0.25 + 0.04 + 0.01 = 0.30. Taking the square root gives 0.547, which is recorded as the deviation distance of node 2. Among the six key nodes in the first cycle, the maximum deviation is calculated to be 0.70, and the minimum deviation is 0.00. A threshold of 0.60 is set as the criterion for judging abnormal nodes. In the first cycle, one node exceeds this threshold. In the calculation of the second cycle, the maximum deviation is 0.85, and the number of nodes exceeding the threshold is 2. Therefore, there are more abnormal nodes in the second cycle. Finally, the frequency index of abnormal deviation nodes is obtained. The frequency in the first cycle is 1 / 6 = 0.167, and the frequency in the second cycle is 2 / 6 = 0.333.

[0038]

[0039] As shown in Table 4, in the second period, the deviation distance of two nodes exceeded the threshold of 0.60, so the frequency of abnormal deviation nodes was higher.

[0040] S413: Based on the frequency index of abnormal deviation nodes, analyze the repetitive deviation trend of spatial trajectory in periodic training, and adjust the corresponding weight of movement standardization in the scoring structure in combination with the frequency of abnormal node occurrence, and output the postoperative rehabilitation score. After obtaining the frequency index of abnormal deviation nodes, it is necessary to analyze the recurring offset trend of the spatial trajectory during periodic training. Specifically, this involves statistically analyzing the changing trend of the abnormal frequency period by period. The frequency in the first period is 0.167, and in the second period it is 0.333, indicating an increase in frequency. A baseline interval is defined as follows. A deviation of 0.20 is considered normal, while a deviation exceeding 0.20 is considered abnormal. Therefore, the first cycle is normal, and the second cycle is abnormal. After obtaining this result, the frequency of abnormal nodes is combined with the rehabilitation scoring structure. The initial weight of movement standardization is set to 0.4. When the abnormal frequency exceeds 0.20, this weight is increased to 0.5. The corresponding adjustment method is: final score = base score × movement standardization weight. Assuming the base score is 80 points, the score for the first cycle with a weight of 0.4 is 80 × 0.4 = 32 points, and the score for the second cycle with a weight of 0.5 is 80 × 0.5 = 40 points. The final output postoperative rehabilitation score is 32 points for the first cycle and 40 points for the second cycle, serving as a stage indicator in the patient's rehabilitation process.

[0041] Please see Figure 6 The specific steps for obtaining patient behavior management records are as follows: S511: Using postoperative rehabilitation scores, analyze the patient's daily training task completion records, action execution frame coverage data and training task interval distribution to obtain the daily actual number of completions, action segment coverage rate and training interruption days, and obtain a set of task completion statistical indicators. Using the postoperative rehabilitation score as input (32 points in cycle 1 and 40 points in cycle 2), we established the baseline data for the rehabilitation score sequence. We then analyzed the patient's daily training task completion records. Assuming the patient planned 10 sets of tasks on day 1, each set containing 20 frames of movement, and actually completed 8 sets, the actual number of completions was 8. We then extracted the frame coverage data, setting each set of movements to cover 20 frames. If the effective frames actually completed in the first set on day 1 were 18, the coverage rate was 18 / 20 = 0.90. We then statistically analyzed the effective frame coverage rate of the 8 sets out of 10, averaging 0.88, resulting in a movement segment coverage rate of 0.88. Next, we analyzed the training task interval distribution, recording training over three consecutive days: 8 sets completed on day 1, 10 sets completed on day 2, and no training on day 3. Therefore, there was one day of interruption, resulting in a training interruption day of 1 day. These values ​​were compiled into a task completion statistical index set, including daily actual completion counts, movement segment coverage rate, and training interruption days, providing a basis for subsequent compliance calculations.

[0042] S512: Based on the task completion statistical indicator set, calculate the percentage of actual completion times, the coverage rate of continuous action segments, and the frequency of training interruption days, construct a compliance indicator sequence, and compare it with the rehabilitation score sequence to obtain compliance fluctuation correlation indicators. Based on the task completion statistics set, it is necessary to calculate the percentage of actual completions, the coverage rate of continuous movement segments, and the frequency of training interruptions to construct a compliance indicator sequence and compare it with the rehabilitation scoring sequence. First, the percentage of actual completions is calculated as the ratio of actual completions to planned completions; for day 1, this is 8 / 10 = 0.80. The coverage rate of movement segments is already calculated as 0.88. The frequency of training interruptions is calculated as the ratio of the number of interrupted days to the total number of days; in this example, 1 / 3 ≈ 0.33. These three indicators are then combined to form the compliance indicator set. Subsequently, a sequence of adherence indicators was constructed, along with a rehabilitation scoring sequence. Correspondingly, the process of calculating the correlation index for adherence fluctuation involves comparing the changes in adherence index with the trends in rehabilitation scores. On day 1, adherence indexes of 0.80, 0.88, and 0.33 correspond to a rehabilitation score of 32. On day 2, adherence indexes of 0.90, 0.92, and 0.0 correspond to a rehabilitation score of 40. The comparison shows that as adherence indexes generally improve, rehabilitation scores also rise, indicating a correlated fluctuation trend. The final correlation index for adherence fluctuation is: .

[0043]

[0044] As shown in Table 5, the compliance indicators improved significantly on both Day 1 and Day 2, and the rehabilitation score improved simultaneously, demonstrating the quantitative correlation between compliance fluctuations and scores.

[0045] S513: Based on the compliance fluctuation correlation index, identify the impact of compliance changes on rehabilitation scores, adjust the patient's rehabilitation plan, and output the patient behavior management record; After obtaining the correlation indicators for adherence fluctuations, it is necessary to identify the impact of adherence changes on rehabilitation scores. The action involved analyzing the differences between adherence indicators and rehabilitation scores daily and labeling them accordingly. On day 1, the average adherence indicator was (0.80 + 0.88 + 0.33) / 3 ≈ 0.67, corresponding to a rehabilitation score of 32. On day 2, the average adherence indicator was (1.00 + 0.92 + 0.00) / 3 ≈ 0.64, corresponding to a rehabilitation score of 40. Comparing the two days revealed little difference in average adherence values, but improvements in the percentage of completed tasks and coverage were more closely related to increases in rehabilitation scores. Therefore, it is necessary to identify... Among the changes in compliance, the percentage of completed sessions and the coverage rate had a more significant impact on score improvement. Based on these results, the rehabilitation plan was adjusted. Specifically, the minimum number of completed sessions was increased from at least 8 out of 10 to at least 9. Simultaneously, the coverage rate assessment standard was increased from 0.85 to 0.90. The resulting patient behavior management record stated that from the third day onwards, at least 9 sessions should be completed daily, with each session achieving a coverage rate of 0.90 or higher to maintain the upward trend in rehabilitation scores. The final patient behavior management record was as follows: .

[0046] A smart assessment system for postoperative rehabilitation of trauma orthopedic patients, the system being used to execute the aforementioned smart assessment method for postoperative rehabilitation of trauma orthopedic patients, the system comprising: The abnormal data identification module calls IoT sensors to collect the angular velocity sequence of the knee and ankle joints during the squatting action, calculates the angular velocity difference sequence in the main rotation direction, compares it with the angular velocity change range of the same node in the previous cycle, analyzes the action mutation rate, detects abnormal monitoring data, and outputs the abnormal data identification results. The muscle group response coordination module calls the abnormal data identification results, combines them with the signals collected by the electromyography sensor, compares the upward trend of the electromyography signal in the initiation phase of the movement with the acceleration direction of the corresponding joint angle change, judges the degree of coordination between the muscle group activation behavior and the joint movement output, and generates response offset information. The muscle group synergy analysis module uses response offset information to compare the activation timing and amplitude ratio of the left and right quadriceps muscles in the alternating leg raise training task. Combined with the direction of change of X-axis posture angle, it analyzes the symmetrical coordination structure between muscle group activation and posture response, identifies structural abnormal movements, and generates muscle group synergy analysis results. The motion trajectory assessment module, based on the results of muscle group synergy analysis, filters a multi-cycle point set of the joint's three-dimensional trajectory during continuous hip abduction training, calculates the deviation from the standard path, determines the frequency of abnormal nodes, adjusts the weight of the movement standardization score, and outputs the postoperative rehabilitation score. The patient behavior management module uses postoperative rehabilitation scores to analyze patients' daily training task completion records, action execution frame coverage, and task interval time distribution, constructs a compliance index sequence, compares it with the rehabilitation score sequence, analyzes the impact of patient behavior fluctuations on functional rehabilitation, adjusts the patient's rehabilitation plan, and outputs patient behavior management records.

[0047] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A smart assessment method for postoperative rehabilitation in trauma orthopedic surgery, characterized in that, Includes the following steps: S1: Call IoT sensors to collect the angular velocity sequence of the knee and ankle joints during the squatting action, calculate the angular velocity difference sequence in the main rotation direction, compare it with the angular velocity change range of the same node in the previous cycle, analyze the action mutation rate, detect abnormal monitoring data, and output the abnormal data identification result; S2: Call the abnormal data identification results, combine them with the signals collected by the electromyography sensor, compare the upward trend of the electromyography signal in the initiation phase of the movement with the acceleration direction of the corresponding joint angle change, determine the degree of coordination between the muscle group activation behavior and the joint movement output response, and generate response offset information. S3: Using the response offset information, compare the activation timing and amplitude ratio of the left and right quadriceps muscles in the alternating leg raise training task, combine the change direction of the X-axis posture angle, analyze the symmetrical coordination structure between muscle group activation and posture response, identify structural abnormal movements, and generate muscle group synergistic analysis results. S4: Based on the muscle group synergy analysis results, screen the multi-cycle point set of the joint three-dimensional trajectory in continuous hip abduction training, calculate the deviation from the standard path, determine the frequency of abnormal nodes, adjust the weight of the movement standardization score, and output the postoperative rehabilitation score.

2. The intelligent assessment method for postoperative rehabilitation in trauma orthopedic surgery according to claim 1, characterized in that, The abnormal data identification results include angular velocity fluctuation amplitude, continuous time mutation points, and abnormal segments in the monitoring sequence. The response offset information specifically includes the direction of change in the upward trend of electromyography, the angular acceleration offset segment, and the proportion of response mismatch time periods. The muscle group coordination analysis results include activation time difference, activation intensity ratio, and posture coordination direction. The postoperative rehabilitation score specifically refers to the trajectory repetition offset frequency, score weight adjustment coefficient, and movement standardization score interval.

3. The intelligent assessment method for postoperative rehabilitation in trauma orthopedic surgery according to claim 1, characterized in that, The specific steps for obtaining the abnormal data identification results are as follows: S111: Call IoT sensors to collect angular velocity data of the knee and ankle joints during continuous squatting movements, extract the angular velocity signal sequence of the main rotation direction in each cycle, arrange the angular velocity point sets in multiple cycles in chronological order, and generate a continuous angular velocity sequence group. S112: Based on the continuous angular velocity sequence group, the angular velocity mutation rate index is calculated by comparing the angular velocity change range of the same node in the previous period, and the angular velocity mutation distribution is obtained by statistically analyzing the fluctuation amplitude of each node. S113: Call the angular velocity mutation distribution, determine the abnormal change trend of the fluctuation amplitude in the time continuous structure, analyze the action mutation ratio, detect abnormal monitoring data, and output the abnormal data identification result.

4. The intelligent assessment method for postoperative rehabilitation in trauma orthopedic surgery according to claim 3, characterized in that, The specific steps for obtaining the response offset information are as follows: S211: Call the abnormal data identification results, collect the signal sequence of the electromyography sensor during the patient's movement initiation phase, fit the slope of the extracted signal rising segment, calculate the electromyography activation trend change rate in each cycle, and extract the acceleration vector by combining the synchronously collected joint angle change information to obtain the electromyography trend response structure. S212: Based on the electromyographic trend response structure, compare the degree of coordination between the activation rate change of the muscle group corresponding to each joint within the same period and the acceleration direction of the target joint, determine the transient deviation structure within the period, calculate the response coordination matching index, and obtain the trend matching fluctuation segment index sequence. S213: Call the trend matching fluctuation segment index sequence to determine the degree of response coordination between muscle group activation behavior and target joint movement output, and identify the proportion of the output response mismatch segment between muscle group nerve activation and movement execution in the time series, and generate response offset information.

5. The intelligent assessment method for postoperative rehabilitation in trauma orthopedic surgery according to claim 4, characterized in that, The specific steps for obtaining the muscle group synergy analysis results are as follows: S311: Based on the response offset information, extract the activation frame indexes of the left and right quadriceps muscles within the alternating leg raise training cycle, compare the time intervals of the activation frames on the left and right sides, determine the degree of activation timing offset, and obtain the activation timing offset index. S312: Based on the activation timing offset index, extract the normalized peak values ​​of electromyography on the left and right sides in each cycle, and calculate the muscle group posture symmetry coordination index by combining the X-axis posture angle, muscle group activation duration, and activation frame position difference of the corresponding cycle. S313: Call the muscle group posture symmetry coordination index, combine it with the activation time sequence offset index, determine the symmetry state of the action execution, determine the periodic structural abnormal action, and obtain the muscle group coordination analysis results.

6. The intelligent assessment method for postoperative rehabilitation in trauma orthopedic surgery according to claim 5, characterized in that, The specific steps for obtaining the postoperative rehabilitation score are as follows: S411: Based on the muscle group synergy analysis results, screen the three-dimensional trajectory point set of the joint in each cycle of the continuous hip abduction training task, extract the key trajectory nodes in each cycle, and obtain the set of key trajectory nodes. S412: Call the set of key nodes of the trajectory, compare the distance between the key nodes of each cycle trajectory and the corresponding points of the standard action path, determine the frequency of occurrence of abnormal deviation nodes in each cycle, and obtain the frequency index of abnormal deviation nodes. S413: Based on the abnormal deviation node frequency index, analyze the repetitive deviation trend of the spatial trajectory in periodic training, and adjust the corresponding weight of the action standardization in the scoring structure in combination with the frequency of abnormal node occurrence, and output the postoperative rehabilitation score.

7. The intelligent assessment method for postoperative rehabilitation in trauma orthopedic surgery according to claim 1, characterized in that, The method further includes: S5: Using the postoperative rehabilitation score, analyze the patient's daily training task completion record, action execution frame coverage, and task interval time distribution, construct a compliance index sequence, compare it with the rehabilitation score sequence, analyze the impact of patient behavior fluctuations on functional rehabilitation, adjust the patient's rehabilitation plan, and output the patient behavior management record. The patient behavior management records include the percentage of training task completion, the percentage of action frame coverage, and the frequency of training interruptions.

8. The intelligent assessment method for postoperative rehabilitation in trauma orthopedic surgery according to claim 7, characterized in that, The specific steps for obtaining the patient behavior management records are as follows: S511: Using the postoperative rehabilitation score, analyze the patient's daily training task completion record, action execution frame coverage data and training task interval distribution to obtain the daily actual number of completions, action segment coverage rate and training interruption days, and obtain a set of task completion statistical indicators. S512: Based on the task completion statistical indicator set, calculate the percentage of actual completion times, the coverage rate of continuous action segments, and the frequency of training interruption days, construct a compliance indicator sequence, and compare it with the rehabilitation score sequence to obtain compliance fluctuation correlation indicators. S513: Based on the aforementioned compliance fluctuation correlation index, identify the impact of compliance changes on rehabilitation scores, adjust the patient's rehabilitation plan, and output patient behavior management records.

9. A smart assessment system for postoperative rehabilitation in trauma orthopedic surgery, characterized in that, The system is used to implement the intelligent assessment method for postoperative rehabilitation in trauma orthopedic surgery as described in any one of claims 1-8, and the system comprises: The abnormal data identification module calls IoT sensors to collect the angular velocity sequence of the knee and ankle joints during the squatting action, calculates the angular velocity difference sequence in the main rotation direction, compares it with the angular velocity change range of the same node in the previous cycle, analyzes the action mutation rate, detects abnormal monitoring data, and outputs the abnormal data identification results. The muscle group response coordination module calls the abnormal data identification results, combines them with the signals collected by the electromyography sensor, compares the upward trend of the electromyography signal in the initiation phase of the movement with the acceleration direction of the corresponding joint angle change, judges the degree of coordination between the muscle group activation behavior and the joint movement output, and generates response offset information. The muscle group synergy analysis module uses the response offset information to compare the activation timing and amplitude ratio of the left and right quadriceps muscles in the alternating leg raise training task. Combined with the change direction of the X-axis posture angle, it analyzes the symmetrical coordination structure between muscle group activation and posture response, identifies structural abnormal movements, and generates muscle group synergy analysis results. Based on the muscle group synergy analysis results, the motion trajectory assessment module filters the multi-cycle point set of the joint three-dimensional trajectory in continuous hip abduction training, calculates the deviation from the standard path, judges the frequency of abnormal nodes, adjusts the weight of the movement standardization score, and outputs the postoperative rehabilitation score. The patient behavior management module uses the postoperative rehabilitation score to analyze the patient's daily training task completion records, action execution frame coverage, and task interval time distribution, constructs a compliance index sequence, compares it with the rehabilitation score sequence, analyzes the impact of patient behavior fluctuations on functional rehabilitation, adjusts the patient's rehabilitation plan, and outputs patient behavior management records.

Images