An intelligent ankle pump motion monitoring system and method

Abstract

This invention discloses an intelligent ankle pump exercise monitoring system and method, belonging to the field of intelligent exercise monitoring technology. Specifically, it includes: real-time acquisition of ankle joint posture and physiological parameters through high-precision sensors and monitoring equipment; obtaining accurate exercise data through data fusion and filtering; setting a standard exercise cycle; comparing and correcting posture sequences with a standard database; assessing exercise standardization, energy consumption, and body load to preliminarily determine body condition; setting body condition grading; classifying risk levels based on posture assessment results; developing personalized optimal posture plans; and real-time monitoring and correction of non-standard postures; and constructing a machine learning model based on historical data to predict exercise duration and lower limb condition, customizing exercise or recovery plans for different groups, thus realizing intelligent and personalized monitoring and management of ankle pump exercises.

Description

Technical Field

[0001] This invention belongs to the field of intelligent motion monitoring technology, specifically an ankle pump motion intelligent monitoring system and method. Background Technology

[0002] Ankle pump exercises are an effective lower limb functional training method. They primarily utilize the rhythmic contractions of the lower limb muscles to move the ankle joint, acting like a pump to promote blood and lymphatic circulation in the lower limbs. This plays a crucial role in post-orthopedic surgery rehabilitation and in preventing deep vein thrombosis (DVT) in the lower limbs. Specifically, ankle pump exercises include ankle flexion, extension, and circumduction movements. These movements effectively promote blood circulation in the lower limbs, preventing lower limb edema, varicose veins, and DVT, while also helping to prevent muscle atrophy and ankle stiffness. However, traditional monitoring methods suffer from insufficient accuracy, poor real-time performance, and inaccurate assessment results. Therefore, a smart ankle pump exercise monitoring system capable of real-time monitoring and precise assessment is needed.

[0003] For example, patent CN107194193B discloses an ankle pump motion monitoring method and device, including: acquiring real-time motion posture information of the user, including angular velocity information, acceleration information, and magnetic force information; fusing the angular velocity information, acceleration information, and magnetic force information to obtain corrected motion posture information; determining the corrected motion posture information sequence within a motion cycle; determining whether the corrected motion posture information sequence matches the standard posture information sequence in the standard posture information sequence database, and obtaining a matching result; sending and / or displaying the matching result to prompt the user whether to continue moving according to the current action. The ankle pump motion monitoring method provided by this technical solution can be implemented with small electronic components, making the monitoring system compact and easy to wear, meeting the actual needs of postoperative patients, and eliminating the need for sophisticated mechanical structures, thus reducing machining costs.

[0004] The existing technologies mentioned above all have the following problems: lack of analysis of the details and depth of the exercise; relatively little personalized assessment and guidance for each target group; certain limitations in predictive ability and support for specific types of exercise; and insufficient risk warning capabilities. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention proposes an intelligent ankle pump exercise monitoring system and method. It utilizes high-precision sensors and monitoring equipment to collect ankle joint posture and physiological parameters in real time, and obtains accurate exercise data through data fusion and filtering. A standard exercise cycle is set, and the posture sequence is compared and corrected with a standard database to assess exercise standardization, energy consumption, and body load, providing a preliminary assessment of the body's condition. A body condition grading system is established, risk levels are classified based on posture assessment results, and personalized optimal posture plans are developed. Incorrect postures are monitored and corrected in real time. A machine learning model is built based on historical data to predict exercise duration and lower limb condition, customizing exercise or recovery plans for different groups. This achieves intelligent and personalized monitoring and management of ankle pump exercises.

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A smart method for monitoring ankle pump movement includes:

[0008] Step S1: Use sensors and monitoring devices to collect ankle joint posture data and physiological parameter information of the target group in real time, and obtain ankle pump movement posture data through data fusion and adaptive filtering methods;

[0009] Step S2: Set a standard exercise cycle. Based on ankle pump movement posture data, use a posture estimation algorithm to extract the corrected movement posture information sequence within the cycle and compare it with the data pre-stored in the standard posture information sequence database to determine whether the ankle pump movement of the target group meets the standard. At the same time, establish a dynamic physiological model to analyze the changing trend of the target group's physiological parameters, assess the target group's physical exertion level and physical load status, and combine posture information and physiological parameters to preliminarily assess the target group's physical condition using a multi-factor comprehensive evaluation algorithm.

[0010] Step S3: Set up a physical condition grading standard. Based on the posture assessment results, judge the exercise quality of the target group. Combined with the physical condition, divide the target group into four risk levels. For the target group with different risk levels, formulate the best posture plan and monitor the exercise posture of the target group in real time. If the target group's posture is found to be non-standard, issue a prompt through the monitoring device to guide them to adjust to the best posture.

[0011] Step S4: Establish a machine learning prediction model. Based on the target group's posture information, physiological parameters, and exercise intensity characteristics, use the machine learning model to predict the time the target group can continue ankle pump exercises and the condition of their lower limbs. Based on the prediction results, develop an exercise plan or recovery program for the target group.

[0012] Specifically, step S2 includes the following steps:

[0013] S2.1: Receive ankle pump movement posture data, set the movement cycle to T, and create a standard posture information sequence database;

[0014] S2.2: Within one motion cycle, the received ankle pump motion posture data are sorted in ascending order according to time sequence to generate a posture information sequence. The dynamic time warping algorithm is used to compare the generated posture information sequence with the standard sequence in the standard posture information sequence database to obtain the sequence difference.

[0015] S2.3: Construct an attitude estimation model and use the attitude estimation model to correct the real-time attitude information sequence based on the sequence difference.

[0016] Specifically, step S2 further includes the following steps:

[0017] S2.4: Based on real-time collected physiological parameters, a dynamic physiological model is established using time series analysis methods, and the parameters of the dynamic physiological model are updated in real time;

[0018] S2.5: Set quantitative indicators for physical exertion and physical load, calculate quantitative indicators based on changes in physiological parameters collected in real time, and assess the level of physical exertion and physical load of the target group based on the quantitative indicator values.

[0019] S2.6: Combining the posture information sequence correction results in step S2.3 and the physical exertion and body load assessment results in step S2.5, assign weights to each factor and use a weighted summation algorithm to calculate the comprehensive assessment result.

[0020] Specifically, the specific steps of the dynamic time warping algorithm in S2.2 include:

[0021] S2.21: Set the initial parameters of the dynamic time warping algorithm and obtain the real-time generated attitude information sequence X = {x1,...,x...} n The standard attitude information sequence Y = {y1,...,y} m}, where x n This represents the nth real-time generated attitude information, where n represents the number of attitude information items in the real-time generated attitude information sequence, y m This represents the m-th standard attitude information, where m represents the number of attitude information in the standard attitude information sequence.

[0022] S2.22: Calculate the point-to-point distance d between any two points in X and Y, and combine the point-to-point distances into a distance matrix D. n×m The formula is:

[0023]

[0024] Where, d A,BThe distance between point A in X and point B in Y represents the point-to-point distance, N represents the number of point pairs in the attitude information, and w k δ represents the weight factor of the point pair in the k-th pose information. k This represents the offset term of the point pair in the k-th attitude information, where p represents a positive real number parameter.

[0025] Specifically, the specific steps of the dynamic time warping algorithm in S2.2 further include:

[0026] S2.23: Based on the distance matrix D n×m Initialize the cumulative distance matrix L n×m And L(1,1)=D(1,1), where L(1,1)=l 1,1 D(1,1) represents the element in the first row and first column of the cumulative distance matrix. 1,1 This represents the element in the first row and first column of the distance matrix;

[0027] S2.24: Calculate the cumulative distance matrix L using an improved dynamic programming algorithm. n×m The formula is:

[0028]

[0029] Among them, l i,j ,i∈n,j∈m represents the element in the i-th row and j-th column of the cumulative distance matrix, d i,j,t W represents the distance to position (i,j) directly at time t. d Cd represents the basic weighting factor. i,j,t Cu i,j,t Cl i,j,t W represents the adjustment terms generated at time t when moving from the diagonal, above, and left respectively to (i,j). diag W up W left These represent the weighting factors when moving diagonally, upwards, and to the left, respectively.

[0030] Specifically, the specific steps of the dynamic time warping algorithm in S2.2 further include:

[0031] S2.25: From the cumulative distance matrix L n×m bottom right corner l n,m Initially, backtracking logic is used to find the shortest path to the point while moving diagonally, upwards, and to the left. This backtracking process is repeated until the point is reached at the top left corner of the cumulative distance matrix. 0,0 ;

[0032] S2.26: Record all points on the path, obtain the best alignment path between sequences X and Y, and use the Euclidean distance formula to obtain the similarity or difference between the real-time attitude information sequence and the standard sequence based on the best alignment path.

[0033] An intelligent ankle pump movement monitoring system includes: a data processing module, a posture assessment module, a monitoring module, and a planning module;

[0034] The data processing module is used to capture ankle joint posture information and physiological parameter information of the target group in real time, and to fuse, filter and correct the collected raw data to obtain ankle pump movement posture data.

[0035] The posture assessment module is used to compare the corrected posture information with a standard database, assess the standard of movement, and analyze the body state in conjunction with physiological parameters.

[0036] The monitoring module is used to provide personalized guidance to the target group based on the assessment results and to monitor their movement posture in real time.

[0037] The planning module is used to predict future athletic performance based on historical data and to tailor training or recovery plans for target groups.

[0038] Specifically, the posture assessment module includes: a posture comparison unit, a physiological parameter analysis unit, and a comprehensive assessment unit;

[0039] The posture comparison unit is used to compare the actual movement posture of the target group with the standard posture and assess the degree of deviation.

[0040] The physiological parameter analysis unit is used to analyze the physiological response of the target group during exercise and to assess the impact of exercise intensity on the body.

[0041] The comprehensive evaluation unit is used to comprehensively evaluate the exercise effect by combining the results of posture comparison and physiological parameter analysis.

[0042] Specifically, the monitoring module includes: a risk classification unit, a plan formulation unit, a real-time monitoring unit, and a guidance unit;

[0043] The risk grading unit is used to assess the risk level of deep vein thrombosis in the lower extremities based on the physiological state and posture information of the target group.

[0044] The program development unit is used to develop or adjust the ankle pump exercise rehabilitation program based on the assessment results.

[0045] The real-time monitoring unit is used to continuously monitor the movement status of the target group and provide real-time feedback and adjustments.

[0046] The guidance unit is used to provide real-time movement guidance to the target group through visual and auditory means.

[0047] Compared with the prior art, the beneficial effects of the present invention are:

[0048] 1. This invention proposes an intelligent ankle pump movement monitoring system and optimizes and improves its architecture, operation steps and processes. The system has the advantages of simple process, low investment and operating costs and low production costs.

[0049] 2. This invention proposes an intelligent ankle pump movement monitoring method. By using sensors and monitoring devices to collect ankle joint posture data and physiological parameter information in real time, it achieves instant capture of the movement status of the target group, which helps to promptly detect abnormal or improper postures during exercise, ensuring the safety and effectiveness of exercise. Combining posture estimation algorithms and a standard posture information sequence database, the method accurately evaluates the ankle pump movement of the target group to determine whether it meets the standards. At the same time, by establishing a dynamic physiological model and analyzing the changing trends of physiological parameters, it can more comprehensively assess the physical exertion and physical load of the target group, which helps to provide customized exercise guidance and suggestions for target groups with different physical conditions and exercise abilities. The method sets physical condition grading standards and risk level classifications to promptly detect potential health risks. For target groups with poor exercise quality or poor physical condition, the system can monitor their movement posture in real time and issue prompts through the monitoring devices when non-standard postures are detected, guiding them to adjust to the optimal posture. This timely intervention helps to prevent sports injuries and protect the health and safety of the target group.

[0050] 3. This invention proposes an intelligent monitoring method for ankle pump exercise, which establishes a machine learning prediction model and uses the postural information, physiological parameters and exercise intensity characteristics of the target group to predict the time they can continue ankle pump exercise and the condition of their lower limbs. This predictive ability helps to develop more scientific and reasonable exercise plans or recovery programs for the target group, avoiding over-exercise or under-exercise. Attached Figure Description

[0051] Figure 1 This is a schematic diagram of an intelligent ankle pump movement monitoring method according to the present invention;

[0052] Figure 2 This is a flowchart illustrating the intelligent monitoring method for ankle pump movement according to the present invention.

[0053] Figure 3 This is a diagram illustrating the architecture of an intelligent ankle pump motion monitoring system according to the present invention. Detailed Implementation

[0054] Example 1

[0055] Please see Figures 1-2The present invention provides an embodiment of an intelligent ankle pump movement monitoring method, comprising the following steps:

[0056] Step S1: Use sensors and monitoring devices to collect ankle joint posture data and physiological parameter information of the target group in real time, and obtain ankle pump movement posture data through data fusion and adaptive filtering methods;

[0057] Furthermore, the specific steps of step S1 include:

[0058] (1) Selection of sensors and monitoring equipment:

[0059] Ankle posture information, including angular velocity, acceleration, and magnetic force, is captured using accelerometers, gyroscopes, and magnetometers. At the same time, physiological parameter monitoring devices, such as heart rate monitors, blood pressure monitors, and blood oxygen saturation monitors, are used to record physiological parameters in real time.

[0060] (2) Sensor arrangement and calibration:

[0061] The sensor is fixed to the ankle joint of the target group to ensure that the sensor can accurately capture the posture changes during the movement. The sensor is calibrated to eliminate initial errors and offsets and ensure data accuracy.

[0062] (3) Data collection:

[0063] The sensors and monitoring equipment are activated to collect ankle joint posture information and physiological parameters in real time, and the collected raw data is initially converted, such as converting acceleration and angular velocity into Euler angles or quaternion posture representations.

[0064] (4) Data fusion:

[0065] In this invention, a complementary filtering data fusion algorithm is used to fuse data from different sensors. Combining the accuracy and noise factors of each sensor, a weighted optimization algorithm is used to obtain attitude estimation.

[0066] (5) Filtering:

[0067] The fused attitude data is low-pass filtered to further eliminate noise and jitter, and the filtered attitude data is output as ankle pump motion attitude data.

[0068] Step S2: Set a standard exercise cycle. Based on ankle pump movement posture data, use a posture estimation algorithm to extract the corrected movement posture information sequence within the cycle and compare it with the data pre-stored in the standard posture information sequence database to determine whether the ankle pump movement of the target group meets the standard. At the same time, establish a dynamic physiological model to analyze the changing trend of the target group's physiological parameters, assess the target group's physical exertion level and physical load status, and combine posture information and physiological parameters to preliminarily assess the target group's physical condition using a multi-factor comprehensive evaluation algorithm.

[0069] Step S3: Set up a physical condition grading standard. Based on the posture assessment results, judge the exercise quality of the target group. Combined with the physical condition, divide the target group into four risk levels. For the target group with different risk levels, formulate the best posture plan and monitor the exercise posture of the target group in real time. If the target group's posture is found to be non-standard, issue a prompt through the monitoring device to guide them to adjust to the best posture.

[0070] The physical condition grading standard includes multiple dimensions such as body mass index, cardiopulmonary function, muscle strength, and flexibility; the optimal posture solution includes but is not limited to correct standing posture, sitting posture, walking posture, and posture control during exercise; at the same time, the physical condition is divided into four levels according to the physical condition grading standard, such as healthy, sub-healthy, mild risk, and high risk.

[0071] In this invention, the posture assessment tool uses a combination of motion capture system and wearable sensors to collect and analyze the posture data of the target group, thereby improving accuracy and real-time performance.

[0072] Step S4: Establish a machine learning prediction model. Based on the target group's posture information, physiological parameters, and exercise intensity characteristics, use the machine learning model to predict the time the target group can continue ankle pump exercises and the condition of their lower limbs. Based on the prediction results, develop an exercise plan or recovery program for the target group.

[0073] Furthermore, the specific steps of step S4 include:

[0074] (1) Receive the postural information of the target group when performing ankle pump exercises, such as joint angles and movement trajectory; physiological parameters, such as heart rate, blood pressure, and blood oxygen saturation; and exercise intensity characteristics, such as exercise frequency and force output, and perform preprocessing.

[0075] (2) Combine posture information, physiological parameters and motion intensity features into a feature set. Each sample contains a complete set of feature data, and the feature set is divided into a training set and a test set.

[0076] (3) Select a regression-classification prediction model based on the characteristics of the problem and the nature of the data, train the regression-classification prediction model using training set data, optimize the performance of the regression-classification prediction model by adjusting model parameters and feature selection, and evaluate and optimize the performance of the regression-classification prediction model using validation set.

[0077] (4) Use test set data to evaluate the final performance of the regression-classification prediction model, including prediction accuracy, recall, and F1 score.

[0078] (5) Predict the newly collected target group data, including the time the target group can continue ankle pump exercises and the condition of the lower limbs;

[0079] (6) Interpret the prediction results, analyze which characteristics have the greatest impact on the prediction results, and develop an exercise plan or recovery program. The program should take into account the physical condition, exercise ability and goals of the target group to ensure that it is both safe and effective.

[0080] The specific steps of step S2 include:

[0081] S2.1: Receive ankle pump movement posture data, set the movement cycle to T, and create a standard posture information sequence database;

[0082] S2.2: Within one motion cycle, the received ankle pump motion posture data are sorted in ascending order according to time sequence to generate a posture information sequence. The dynamic time warping algorithm is used to compare the generated posture information sequence with the standard sequence in the standard posture information sequence database to obtain the sequence difference.

[0083] S2.3: Construct an attitude estimation model and use the attitude estimation model to correct the real-time attitude information sequence based on the sequence difference.

[0084] This invention selects YOLOv7 as the base model for the pose estimation model, trains the base model using a labeled dataset, and optimizes the base model parameters through the backpropagation algorithm, so that the base model can accurately predict human pose and obtain the pose estimation model.

[0085] S2.4: Based on real-time collected physiological parameters, a dynamic physiological model is established using time series analysis methods, and the parameters of the dynamic physiological model are updated in real time. The time series analysis method is existing technology in this field and is not an inventive solution of this application, so it will not be described in detail here.

[0086] S2.5: Set quantitative indicators for physical exertion and physical load, calculate quantitative indicators based on changes in physiological parameters collected in real time, and assess the level of physical exertion and physical load of the target group based on the quantitative indicator values.

[0087] Furthermore, the specific steps of S2.5 include:

[0088] (1) Define quantitative indicators:

[0089] Key physiological parameters for measuring physical exertion and physical load, such as heart rate, blood lactate concentration, maximum oxygen uptake, exercise duration, and energy expenditure, are determined. Based on sports science theory and practical experience, thresholds or ranges for these physiological parameters are set to differentiate between different levels of physical exertion and physical load.

[0090] (2) Data Acquisition:

[0091] Wearable devices, such as heart rate monitors, blood lactate analyzers, and sports watches, are used to collect physiological parameter data of the target group in real time, ensuring the accuracy and timeliness of data collection so as to reflect changes in physical exertion and physical load in a timely manner.

[0092] (3) Data processing:

[0093] The collected physiological parameter data are cleaned and organized to remove outliers and noise. Based on preset quantitative indicators and algorithms, the physical exertion and physical load levels of the target group are calculated.

[0094] (4) Calculation of quantitative indicators:

[0095] Convert physiological parameter data into quantitative indicators, for example, use heart rate reserve percentage to calculate exercise intensity, or use training impulse model to quantify training load;

[0096] (5) Evaluation and feedback:

[0097] Based on quantitative indicators, assess the physical exertion and load levels of the target group to determine whether they are within an appropriate range, and provide feedback on the assessment results to staff such as coaches, athletes, and health consultants so that they can adjust training plans or provide health advice based on the assessment results.

[0098] S2.6: Combining the posture information sequence correction results in step S2.3 and the physical exertion and body load assessment results in step S2.5, assign weights to each factor and use a weighted summation algorithm to calculate the comprehensive assessment result.

[0099] The specific steps of the dynamic time warping algorithm in S2.2 include:

[0100] S2.21: Set the initial parameters of the dynamic time warping algorithm and obtain the real-time generated attitude information sequence X = {x1,...,x...} n The standard attitude information sequence Y = {y1,...,y} m}, where x nThis represents the nth real-time generated attitude information, where n represents the number of attitude information items in the real-time generated attitude information sequence, y m This represents the m-th standard attitude information, where m represents the number of attitude information in the standard attitude information sequence.

[0101] S2.22: Calculate the point-to-point distance d between any two points in X and Y, and combine the point-to-point distances into a distance matrix D. n×m ,in, The formula is:

[0102]

[0103] Where, d A,B The distance between point A in X and point B in Y represents the point-to-point distance, N represents the number of point pairs in the attitude information, and w k δ represents the weight factor of the point pair in the k-th pose information. k Let δ represent the offset term of the point pair in the k-th attitude information, where p represents a positive real number parameter. When p = 1, it is the weighted Manhattan distance; when p = 2, the offset term is not considered, i.e., δ k =0 is the weighted Euclidean distance, and when p→∞, without considering the offset term, it is the weighted Chebyshev distance;

[0104] S2.23: Based on the distance matrix D n×m Initialize the cumulative distance matrix L n×m And L(1,1)=D(1,1), where L(1,1)=l 1,1 D(1,1) represents the element in the first row and first column of the cumulative distance matrix. 1,1 This represents the element in the first row and first column of the distance matrix;

[0105] S2.24: Calculate the cumulative distance matrix L using an improved dynamic programming algorithm. n×m The formula is:

[0106]

[0107] Among them, l i,j ,i∈n,j∈m represents the element in the i-th row and j-th column of the cumulative distance matrix, d i,j,t W represents the distance to position (i,j) directly at time t. d Cd represents the basic weighting factor. i,j,t Cu i,j,t Cl i,j,t W represents the adjustment terms generated at time t when moving from the diagonal, above, and left respectively to (i,j). diag W up W leftThese represent the weighting factors when moving diagonally, upwards, and to the left, respectively.

[0108] S2.25: From the cumulative distance matrix L n×m bottom right corner l n,m Initially, backtracking logic is used to find the shortest path to the point while moving diagonally, upwards, and to the left. This backtracking process is repeated until the point is reached at the top left corner of the cumulative distance matrix. 0,0 ;

[0109] The backtracking logic includes:

[0110] (1) From l n,m To begin, you need to decide whether to move left, up, or to the upper left to find the optimal predecessor to achieve this score. This usually depends on the scoring rules, such as whether affine gap penalties are used or whether the cost of character matching / non-matching is considered.

[0111] (a) Move to the left l n,m-1 This means that a gap has been inserted in sequence Y, or that the current character of sequence Y has been compared with a "gap" in sequence X.

[0112] (b) Move upward l n-1,m This means that a gap has been inserted in sequence X, or that the current character of sequence X has been compared with a "gap" in sequence Y.

[0113] (c) Move to the upper left l n-1,m-1 This means that the current characters of sequences X and Y have been compared, i.e., matched or not matched;

[0114] (2) Select the one with the highest score among the three directions as the next step.

[0115] S2.26: Record all points on the path, obtain the best alignment path between sequences X and Y, and use the Euclidean distance formula to obtain the similarity or difference between the real-time attitude information sequence and the standard sequence based on the best alignment path.

[0116] Furthermore, the sequence difference is defined as the similarity or change in pose information between adjacent frames. The similarity is evaluated by calculating the distance and angle parameters between point pairs, while the change is obtained by calculating the difference in position between point pairs between adjacent frames. The similarity calculation formula is prior art in this field and is not an inventive solution of this application, so it will not be elaborated here.

[0117] Example 2

[0118] Please see Figure 3 Another embodiment of the present invention provides: an intelligent ankle pump movement monitoring system, comprising:

[0119] Data processing module, attitude assessment module, monitoring module, and planning module;

[0120] The data processing module is used to capture ankle joint posture information and physiological parameter information of the target group in real time, and to fuse, filter and correct the collected raw data to obtain ankle pump movement posture data.

[0121] The posture assessment module is used to compare the corrected posture information with a standard database, assess the standard of movement, and analyze the body state in conjunction with physiological parameters.

[0122] The monitoring module is used to provide personalized guidance to the target group based on the assessment results and to monitor their movement posture in real time;

[0123] The planning module is used to predict future athletic performance based on historical data and to tailor training or recovery plans for target groups.

[0124] The data processing module includes: a data acquisition unit, a data fusion unit, and an attitude correction unit;

[0125] The data acquisition unit is used to collect ankle joint posture information, including movement angle, speed, and force.

[0126] The data fusion unit is used to integrate ankle joint posture information and physiological parameter information from different sensors;

[0127] The posture correction unit is used to analyze and determine whether the movement posture of the target group is standard based on the collected posture information and physiological parameter information through a posture estimation algorithm.

[0128] The posture assessment module includes: a posture comparison unit, a physiological parameter analysis unit, and a comprehensive assessment unit;

[0129] The posture comparison unit is used to compare the actual movement posture of the target group with the standard posture and assess the degree of deviation.

[0130] The physiological parameter analysis unit is used to analyze the physiological responses of the target group during exercise, such as heart rate and blood oxygen saturation, and to assess the impact of exercise intensity on the body.

[0131] The comprehensive assessment unit combines the results of posture comparison and physiological parameter analysis to provide a comprehensive assessment of exercise effectiveness, including exercise quality and potential risks.

[0132] The monitoring module includes: a risk classification unit, a plan development unit, a real-time monitoring unit, and a guidance unit;

[0133] The risk grading unit is used to assess the risk level of deep vein thrombosis in the lower extremities based on the physiological status and posture information of the target group, so as to facilitate timely preventive measures.

[0134] The program development unit is used to develop or adjust the ankle pump exercise rehabilitation program based on the assessment results, including exercise intensity, frequency, and duration.

[0135] The real-time monitoring unit is used to continuously monitor the movement status of the target group, ensuring real-time data feedback during the exercise process, which facilitates timely adjustments.

[0136] The guidance unit is used to provide real-time exercise guidance to the target group through visual and auditory means, such as adjusting exercise speed and reminding them to rest.

[0137] The planning module includes: a forecasting unit, a results analysis unit, and a planning unit.

[0138] Prediction unit, used to predict the rehabilitation progress of the target group based on historical data and current conditions;

[0139] The results analysis unit is used to analyze the exercise data and rehabilitation effects of the target group;

[0140] The planning unit is used to develop long-term rehabilitation training plans for the target group based on forecast and analysis results, set phased goals, and adjust strategies.

[0141] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0142] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0143] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments under the guidance of the present invention without departing from the spirit and scope of the present invention. All of these variations are within the protection scope of the present invention.

Claims

1. An ankle pump exercise intelligent monitoring method, characterized in that, include: Step S1: Use sensors and monitoring devices to collect ankle joint posture data and physiological parameter information of the target group in real time, and obtain ankle pump movement posture data through data fusion and adaptive filtering methods; Step S2: Set a standard exercise cycle. Based on ankle pump movement posture data, use a posture estimation algorithm to extract the corrected movement posture information sequence within the cycle and compare it with the data pre-stored in the standard posture information sequence database to determine whether the ankle pump movement of the target group meets the standard. At the same time, establish a dynamic physiological model to analyze the changing trend of the target group's physiological parameters, assess the target group's physical exertion level and physical load status, and combine posture information and physiological parameters to preliminarily assess the target group's physical condition using a multi-factor comprehensive evaluation algorithm. Step S3: Set up a physical condition grading standard. Based on the posture assessment results, judge the exercise quality of the target group. Combined with the physical condition, divide the target group into four risk levels. For the target group with different risk levels, formulate the best posture plan and monitor the exercise posture of the target group in real time. If the target group's posture is found to be non-standard, issue a prompt through the monitoring device to guide them to adjust to the best posture. Step S4: Establish a machine learning prediction model. Based on the target group's posture information, physiological parameters, and exercise intensity characteristics, use the machine learning model to predict the time the target group can continue ankle pump exercises and the condition of their lower limbs. Based on the prediction results, develop an exercise plan or recovery program for the target group. The specific steps of step S2 include: S2.1: Receive ankle pump movement posture data, set the movement cycle to T, and create a standard posture information sequence database; S2.2: Within one motion cycle, the received ankle pump motion posture data are sorted in ascending order according to time sequence to generate a posture information sequence. The dynamic time warping algorithm is used to compare the generated posture information sequence with the standard sequence in the standard posture information sequence database to obtain the sequence difference. S2.3: Construct an attitude estimation model and use the attitude estimation model to correct the real-time attitude information sequence based on the sequence difference. The specific steps of the dynamic time warping algorithm in S2.2 include: S2.21: set initial parameters of a dynamic time warping algorithm, obtain a real-time generated posture information sequence , a standard posture information sequence , wherein, represents the nth real-time generated posture information, n represents the number of posture information in the real-time generated posture information sequence, represents the mth standard posture information, m represents the number of posture information in the standard posture information sequence; S2.22: Calculate the point pair distance d for any point in X and Y and combine the point pair distances into a distance matrix , the formula is: ； in, N represents the point-to-point distance between point A in X and point B in Y, and N represents the number of point pairs in the attitude information. This represents the weight factor of the point pair in the k-th pose information. This represents the offset term of the point pair in the k-th attitude information, where p represents a positive real number parameter; The specific steps of the dynamic time warping algorithm in S2.2 also include: S2.23: Based on the distance matrix Initialize the cumulative distance matrix ,and ,in, This represents the element in the first row and first column of the cumulative distance matrix. This represents the element in the first row and first column of the distance matrix; S2.24: Calculate the cumulative distance matrix using an improved dynamic programming algorithm. The formula is: ； in, This represents the element in the i-th row and j-th column of the cumulative distance matrix. This indicates that the location is reached directly at time point t. distance, Indicates the basic weighting factor. , , These represent the distances taken at time t from the diagonal, above, and left sides, respectively. Adjustments generated at that time , , These represent the weighting factors when moving diagonally, upwards, and to the left, respectively. The specific steps of the dynamic time warping algorithm in S2.2 also include: S2.25: From the cumulative distance matrix bottom right corner Initially, backtracking logic is used to find the shortest path to the point while moving diagonally, upwards, and to the left. This backtracking process is repeated until the top left corner of the cumulative distance matrix is ​​reached. ; S2.26: Record all points on the path, obtain the best alignment path between sequences X and Y, and use the Euclidean distance formula to obtain the similarity or difference between the real-time attitude information sequence and the standard sequence based on the best alignment path.

2. The ankle pump movement intelligent monitoring method as described in claim 1, characterized in that, The specific steps of step S2 also include: S2.4: Based on real-time collected physiological parameters, a dynamic physiological model is established using time series analysis methods, and the parameters of the dynamic physiological model are updated in real time; S2.5: Set quantitative indicators for physical exertion and physical load, calculate quantitative indicators based on changes in physiological parameters collected in real time, and assess the level of physical exertion and physical load of the target group based on the quantitative indicator values. S2.6: Combining the posture information sequence correction results in step S2.3 and the physical exertion and body load assessment results in step S2.5, assign weights to each factor and use a weighted summation algorithm to calculate the comprehensive assessment result.

3. An ankle pump movement intelligent monitoring system, used to implement the ankle pump movement intelligent monitoring method according to any one of claims 1-2, characterized in that, include: Data processing module, attitude assessment module, monitoring module, and planning module; The data processing module is used to capture ankle joint posture information and physiological parameter information of the target group in real time, and to fuse, filter and correct the collected raw data to obtain ankle pump movement posture data. The posture assessment module is used to compare the corrected posture information with a standard database, assess the standard of movement, and analyze the body state in conjunction with physiological parameters. The monitoring module is used to provide personalized guidance to the target group based on the assessment results and to monitor their movement posture in real time. The planning module is used to predict future athletic performance based on historical data and to tailor training or recovery plans for target groups.

4. The ankle pump motion intelligent monitoring system as described in claim 3, characterized in that, The posture assessment module includes: a posture comparison unit, a physiological parameter analysis unit, and a comprehensive assessment unit; The posture comparison unit is used to compare the actual movement posture of the target group with the standard posture and assess the degree of deviation. The physiological parameter analysis unit is used to analyze the physiological response of the target group during exercise and to assess the impact of exercise intensity on the body. The comprehensive evaluation unit is used to comprehensively evaluate the exercise effect by combining the results of posture comparison and physiological parameter analysis.

5. The ankle pump motion intelligent monitoring system as described in claim 4, characterized in that, The monitoring module includes: a risk classification unit, a plan formulation unit, a real-time monitoring unit, and a guidance unit; The risk grading unit is used to assess the risk level of deep vein thrombosis in the lower extremities based on the physiological state and posture information of the target group. The program development unit is used to develop or adjust the ankle pump exercise rehabilitation program based on the assessment results. The real-time monitoring unit is used to continuously monitor the movement status of the target group and provide real-time feedback and adjustments. The guidance unit is used to provide real-time movement guidance to the target group through visual and auditory means.

Images