A rehabilitation training evaluation method and system for elbow joint

Abstract

The present invention relates to the fields of rehabilitation medicine and biomedical engineering, and in particular to a method and system for evaluating rehabilitation training of elbow joints. A method for evaluating rehabilitation training of elbow joints, characterized in that it comprises the following steps: S1: obtaining the examination information of rehabilitation personnel, and conducting a preliminary evaluation of rehabilitation training of the rehabilitation personnel's elbow joints; S2: obtaining the electrode usage of the rehabilitation personnel, and conducting a secondary evaluation based on the electrode detection data; S3: refining the rehabilitation personnel's evaluation report, and adjusting the rehabilitation training plan based on the rehabilitation personnel's evaluation report; S4: based on the adjusted rehabilitation training plan, re-evaluating the degree of recovery of the rehabilitation personnel's elbow joints; S5: obtaining a final evaluation report based on the degree of recovery of the rehabilitation personnel's elbow joints. The present invention dynamically adjusts the rehabilitation plan by quantifying the degree of dependence and muscle synergy, improves the accuracy of the evaluation, ensures that the rehabilitation process is more personalized and efficient, and helps patients recover faster.

Description

Technical Field

[0001] The present invention relates to the fields of rehabilitation medicine and biomedical engineering, and in particular to a rehabilitation training evaluation method and system for an elbow joint. Background Art

[0002] Elbow rehabilitation training is essential for restoring joint function, enhancing muscle strength, improving flexibility, and preventing re-injury. Effective assessment methods can not only ensure the effectiveness and safety of training, but also accelerate the recovery process and avoid injuries caused by improper training. Current assessment methods mainly include ROM measurement, muscle strength testing, and specific activity simulation, which focus on evaluating the recovery of elbow injuries. However, existing methods are highly subjective, lack standardized assessment tools, and ignore individual differences in patients. In order to solve these problems, we need to develop more objective standards and tools, establish a standardized assessment process, and pay attention to individual differences to improve the accuracy of the assessment, better guide rehabilitation training, and promote the recovery process of patients. Summary of the invention

[0003] In order to overcome the shortcoming of low accuracy in evaluation of rehabilitation training of elbow joints, the present invention provides a method and system for evaluating rehabilitation training of elbow joints.

[0004] The technical implementation scheme of the present invention is: a rehabilitation training evaluation method for an elbow joint, comprising the following steps:

[0005] S1: Obtain the rehabilitation personnel's examination information and conduct a preliminary assessment of the rehabilitation personnel's elbow joint rehabilitation training;

[0006] S2: Obtain the electrode usage of rehabilitation personnel and conduct secondary evaluation based on the electrode detection data;

[0007] S3: Refine the rehabilitation personnel assessment report and adjust the rehabilitation training plan according to the rehabilitation personnel assessment report;

[0008] S4: According to the adjusted rehabilitation training plan, re-evaluate the degree of recovery of the elbow joint of the rehabilitation personnel;

[0009] S5: Obtain the final assessment report based on the rehabilitation personnel’s elbow recovery degree.

[0010] Preferably, the obtaining of examination information of rehabilitation personnel and conducting preliminary assessment of rehabilitation training of the rehabilitation personnel's elbow joint include: collecting rehabilitation personnel's elbow joint injury type data, physical characteristic data, and recovery type data, constructing an LSTM model, inputting the collected data into the LSTM model for training, and using the trained LSTM model to obtain preliminary assessment results of the rehabilitation training of the rehabilitation personnel's elbow joint.

[0011] Preferably, the method of obtaining the electrode usage of rehabilitation personnel and performing a secondary evaluation based on the electrode detection data includes: collecting the electrode usage of rehabilitation personnel; extracting electrode stimulation data and electrode detection data from the electrode usage data; constructing a dependency model using the electrode stimulation data; constructing a muscle synergy model using the electrode detection data; constructing a secondary evaluation model combining the dependency model and the muscle synergy model; and obtaining a secondary evaluation report using the secondary evaluation model.

[0012] Preferably, the construction of a dependency model based on electrode stimulation data includes: extracting electrode stimulation therapy effectiveness data and electrode stimulation therapy frequency data; calculating a first dependency of rehabilitation personnel on electrode stimulation; evaluating the degree of negative impact of the rehabilitation process and the imbalance of the rehabilitation plan; obtaining the length of stay in a certain stage of the rehabilitation stage; calculating changes in the degree of electrode stimulation dependence of rehabilitation personnel to obtain a second dependency; and constructing a dependency model by combining the first dependency and the second dependency.

[0013] Preferably, the method of constructing a dependency model based on the first dependency and the second dependency includes: extracting muscle synergy data in the muscle synergy model; obtaining muscle synergy coordination data under the first dependency and muscle synergy coordination data under the second dependency, defining the muscle synergy coordination data under the first dependency as first positive coordination data, and defining the muscle synergy coordination data under the second dependency as first reverse coordination data; and constructing a dependency model based on the first positive coordination data and the first reverse coordination data.

[0014] Preferably, the construction of the dependency model based on the first forward coordination data and the first reverse coordination data includes: extracting muscle coordination data in the muscle coordination model, including coordination of muscles in pairs, in pairs, and up to the overall coordination data of the muscles at the entire elbow joint; constructing a transformer model according to the rehabilitation stage of the rehabilitation personnel, and inputting the muscle coordination data training set into the transformer model for training; obtaining the adjustment angle of the elbow joint under muscle coordination; extracting muscle coordination data that meets the elbow joint adjustment angle of the rehabilitation personnel in the current rehabilitation stage; constructing a neural network model, the input data is the muscle coordination data that meets the elbow joint adjustment angle of the rehabilitation personnel in the current rehabilitation stage, calculating the first forward coordination data and the first reverse coordination data to obtain the offset factor, and obtaining the muscle coordination data that cannot be offset by the offset factor within the stage time; analyzing the muscle coordination data that passes the offset factor to obtain an evaluation report of the rehabilitation personnel;

[0015] The specific calculation formula of the offset factor is:

[0016]

[0017] , where the offset factor is the screening degree value used to promote or slow down the muscle coordination data in the neural network model, the first positive coordination data: represents the data that plays a positive role in muscle synergy, the first reverse coordination data: represents the data that plays a negative role in muscle synergy; the formula calculates the ratio of the difference between the first positive coordination data and the first reverse coordination data to the sum of the two, and its value range is between -1 and 1; a positive value indicates that the positive coordination is dominant, a negative value indicates that the reverse coordination is dominant, and a value close to 0 indicates that the two coordinations roughly offset each other; where the positive coordination data in the formula also includes calculating the proportional relationship between the first positive coordination data and the first reverse coordination data, that is, whether the two satisfy a certain matching relationship,

[0018] First forward matching data = k × first reverse matching data + q

[0019] , where k is the proportional factor and q is the parameter. The formula calculates the proportional relationship between the first positive matching data and the first negative matching data. When the first positive matching data is proportional to the first negative matching data, the original value of the first positive matching data of the offset factor is changed to change the range of the offset factor. The specific calculation method of the first positive matching data in the offset factor is:

[0020] The first positive coordination data = Z × the first reverse coordination data, wherein the first reverse coordination data is normalized to -1 < first reverse coordination data < 0, the role of Z is to ensure that the absolute value of the first positive coordination data is greater than the absolute value of the first reverse coordination data, Z is a minimum value greater than the integer 1, so the range of the offset factor is from negative infinity to 0; wherein, in the neural network model, firstly, the input data of the muscle coordination data is recorded, and at the same time, the muscle coordination data passing through the offset factor is recorded, and then the fuzzy reasoning model is constructed, and finally the rehabilitation training evaluation of the rehabilitation personnel is carried out, and the formula is

[0021]

[0022] , where P is the evaluation score, q, q1, and q2 are three different weights, and the effective ratios of the coordinated muscle data and the muscle coordination data calculated by the fuzzy reasoning model through the offset factors.

[0023] Preferably, the refining of the rehabilitation personnel assessment report and adjusting the rehabilitation training plan according to the rehabilitation personnel assessment report include: extracting elbow joint rehabilitation training data according to the rehabilitation personnel assessment report, and collecting lateral limb muscle detection data according to the corresponding relationship; selecting a dynamic adjustment lateral limb recovery plan according to the rehabilitation personnel's rehabilitation plan; obtaining offset factor data, and when the offset factor data is a positive value and the first positive coordination data is not proportional to the first negative coordination data, constructing a double limb balance model and an RNN model; inputting the muscle coordination data training set of the injured elbow joint limb and the uninjured elbow joint limb into the double limb balance model, dynamically adjusting the training plan of the injured elbow joint limb according to the output result of the double limb balance model, and performing elbow joint training evaluation, inputting the rehabilitation process data training set of the rehabilitation personnel into the RNN model, obtaining the predicted recovery degree value, and performing normalization processing as the weight coefficient; the formula used in the double limb balance model includes,

[0024]

[0025] , where k1 and k3 are set according to the rehabilitation goals or physical characteristics of the rehabilitation personnel themselves, and k2 is the weight coefficient obtained by normalizing the recovery degree value predicted by the RNN model; in actual applications, the values ​​of k1 and k3 are dynamically adjusted according to the rehabilitation goals or physical characteristics of the rehabilitation personnel themselves.

[0026] Preferably, the degree of recovery of the elbow joint of the rehabilitated person is re-evaluated according to the adjusted rehabilitation training plan, including: executing the adjusted rehabilitation training plan to ensure that each link of the rehabilitation training plan is effectively implemented; regularly evaluating the recovery of the elbow joint of the rehabilitated person, comprehensively evaluating the rehabilitation effect through objective indicators and subjective feelings, ensuring the effectiveness of the rehabilitation training plan, and making necessary adjustments based on the evaluation results.

[0027] Preferably, the final evaluation report is obtained according to the degree of recovery of the elbow joint of the rehabilitated person, including: conducting a comprehensive and detailed comprehensive evaluation of the rehabilitation effect according to the actual degree of recovery of the elbow joint of the rehabilitated person, and the report records in detail the rehabilitation progress, existing problems and targeted follow-up suggestion information.

[0028] A rehabilitation training evaluation system for an elbow joint, comprising:

[0029] Data collection module: used to obtain data on elbow injury type, physical characteristics, recovery type, and electrode usage of rehabilitation personnel;

[0030] Evaluation and adjustment module: used for LSTM model construction, dependency model construction, muscle synergy model construction, secondary evaluation model construction, data analysis, plan adjustment, and dynamic regulation;

[0031] Assessment execution and report generation module: used for rehabilitation training execution, assessment, comprehensive assessment, and report generation.

[0032] Beneficial effects: First, the basic information of the rehabilitation personnel is obtained and the LSTM model is constructed for preliminary evaluation to ensure the accuracy of the evaluation. Subsequently, the dependency model and muscle coordination model are constructed by collecting the electrode usage and the data generated, and the evaluation results are further refined. On this basis, a new evaluation model is proposed, which can comprehensively consider the positive and negative effects of electrode stimulation to ensure the comprehensiveness of the evaluation. In particular, the concept of "offset factor" is introduced, which can quantify the influence of positive and negative muscle coordination data. When the first positive coordination data is proportional to the first negative coordination data, it indicates that the patient has become dependent on the electrode stimulation. At this time, the offset factor should approach negative infinity, so as to exclude those data that show good muscle coordination but are actually not conducive to rehabilitation. In order to achieve this goal, the first negative coordination data is normalized to between -1 and 0, and a minimum value Z greater than 1 is used to ensure that the absolute value of the first positive coordination data is always greater than the absolute value of the first negative coordination data, so as to ensure that the range of the offset factor can be extended from negative infinity to 0. Next, the neural network model and fuzzy reasoning model are used to further optimize the evaluation process to achieve more accurate evaluation results. In particular, when constructing the double-limb balance model, the personalized needs of different individuals were taken into account, and the model parameters k1 and k3 were dynamically adjusted to adapt to the specific conditions of the rehabilitation personnel. Finally, the rehabilitation progress was predicted by the RNN model and used as a weight coefficient to further optimize the rehabilitation training plan. The entire evaluation process not only focused on the effect of rehabilitation training, but also ensured the accuracy and practicality of the evaluation results, providing rehabilitation personnel with personalized rehabilitation plans and improving the accuracy of elbow rehabilitation training evaluation. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 The present invention is a flowchart of a method and system for evaluating rehabilitation training of an elbow joint.

[0034] Figure 2 The present invention is a flow chart of constructing a dependency model according to first forward matching data and first reverse matching data.

[0035] Figure 3 The present invention is a schematic structural diagram of a rehabilitation training evaluation system for an elbow joint. DETAILED DESCRIPTION

[0036] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0037] Embodiment 1: A method for evaluating rehabilitation training of an elbow joint, such as Figure 1 As shown, the following steps are included:

[0038] S1: Obtain the rehabilitation personnel's examination information and conduct a preliminary assessment of the rehabilitation personnel's elbow joint rehabilitation training;

[0039] S2: Obtain the electrode usage of rehabilitation personnel and conduct secondary evaluation based on the electrode detection data;

[0040] S3: Refine the rehabilitation personnel assessment report and adjust the rehabilitation training plan according to the rehabilitation personnel assessment report;

[0041] S4: According to the adjusted rehabilitation training plan, re-evaluate the degree of recovery of the elbow joint of the rehabilitation personnel;

[0042] S5: Obtain the final assessment report based on the rehabilitation personnel’s elbow recovery degree.

[0043] Obtain the examination information of rehabilitation personnel and conduct a preliminary assessment of the rehabilitation training of the rehabilitation personnel's elbow joint, including: collecting the rehabilitation personnel's elbow joint injury type data, physical characteristic data, and recovery type data, building an LSTM model, inputting the collected data into the LSTM model for training, and using the trained LSTM model to obtain the preliminary assessment results of the rehabilitation training of the rehabilitation personnel's elbow joint.

[0044] Obtain the electrode usage of rehabilitation personnel and conduct secondary evaluation based on the electrode detection data, including: collecting the electrode usage of rehabilitation personnel; extracting electrode stimulation data and electrode detection data from the electrode usage data; using the electrode stimulation data to build a dependency model; using the electrode detection data to build a muscle synergy model; combining the dependency model with the muscle synergy model to build a secondary evaluation model; using the secondary evaluation model to obtain a secondary evaluation report.

[0045] A dependency model is constructed based on electrode stimulation data, including: extracting electrode stimulation treatment effectiveness data and electrode stimulation treatment frequency data; calculating the first dependency of rehabilitation personnel on electrode stimulation; evaluating the negative impact of the rehabilitation process and the imbalance of the rehabilitation plan; obtaining the length of stay in a certain stage of the rehabilitation stage; calculating the changes in the degree of electrode stimulation dependence of rehabilitation personnel to obtain the second dependency; and constructing a dependency model by combining the first dependency and the second dependency.

[0046] Among them, the first degree of dependence refers to the effectiveness of electrode stimulation treatment (pain relief) which may cause patients to become dependent on it; the second degree of dependence means that if patients become dependent, it will lead to a negative impact on rehabilitation, and the negative impact of the rehabilitation process will lead to an imbalance in the rehabilitation plan. The imbalance in the rehabilitation plan will cause the rehabilitation personnel to stay in a certain rehabilitation stage for a long time. Among them, the imbalance in the rehabilitation plan and the rehabilitation personnel staying in a certain stage for a long time will lead to the patient's dependence. The first and second degrees of dependence calculated at this time are only an analysis of the electrode stimulation data, and do not represent the degree of dependence of the rehabilitation personnel.

[0047] A dependency model is constructed according to the first dependency and the second dependency, including: extracting muscle synergy data in the muscle synergy model; obtaining muscle synergy data under the first dependency and muscle synergy data under the second dependency, defining the muscle synergy data under the first dependency as first positive coordination data, and defining the muscle synergy data under the second dependency as first reverse coordination data; and constructing the dependency model according to the first positive coordination data and the first reverse coordination data.

[0048] Through the analysis of electrode stimulation data, the dependence degree is introduced, but the dependence degree at this time cannot reflect the degree of dependence of rehabilitation personnel on electrode stimulation, let alone the positive or negative dependence of rehabilitation personnel on electrode stimulation. Therefore, by introducing muscle coordination data, the muscle coordination data is analyzed under the dependence degree, and the muscle synergy cooperation data under the first dependence degree is defined as the first positive cooperation data. This is because the effectiveness of electrode stimulation treatment may have positive or negative effects on rehabilitation personnel, but in the rehabilitation process of rehabilitation personnel, the rehabilitation degree of rehabilitation personnel generally recovers to a positive direction. Therefore, the muscle synergy cooperation data under the first dependence degree is defined as the first positive cooperation data; at the same time, the muscle synergy cooperation data under the second dependence degree is defined as the first negative cooperation data. The data is because if the rehabilitation personnel become dependent on electrode stimulation, it will lead to a negative impact on the rehabilitation process, the negative impact on the rehabilitation process will lead to an imbalance in the rehabilitation plan, and the imbalance in the rehabilitation plan will cause the rehabilitation personnel to stay in a certain stage for a long time. Therefore, during the rehabilitation process of the rehabilitation personnel, the muscle synergy data under the second dependence degree will have a negative impact on the rehabilitation personnel, so the muscle synergy data under the second dependence degree is defined as the first reverse coordination data; at this time, the first positive coordination data and the first reverse coordination data are only an analysis of the rehabilitation personnel's dependence on electrode stimulation, and do not indicate the necessity or certainty of the rehabilitation personnel's dependence on electrode stimulation (that is, the first positive coordination data has a favorable effect on the rehabilitation personnel, and the first reverse coordination data has a negative effect on the rehabilitation personnel).

[0049] A dependency model is constructed according to the first forward coordination data and the first reverse coordination data, including: extracting muscle coordination data in the muscle coordination model, including muscle coordination in pairs, three-to-three, and even the overall coordination data of the muscles at the entire elbow joint; constructing a transformer model according to the rehabilitation stage of the rehabilitation personnel, and inputting the muscle coordination data training set into the transformer model for training; obtaining the adjustment angle of the elbow joint under muscle coordination; extracting muscle coordination data that meets the elbow joint adjustment angle of the rehabilitation personnel in the current rehabilitation stage; constructing a neural network model, the input data is the muscle coordination data that meets the elbow joint adjustment angle of the rehabilitation personnel in the current rehabilitation stage, calculating the first forward coordination data and the first reverse coordination data to obtain the offset factor, and obtaining the muscle coordination data that cannot be offset by the offset factor within the stage time; analyzing the muscle coordination data that passes the offset factor to obtain an evaluation report of the rehabilitation personnel;

[0050] The specific calculation formula of the offset factor is:

[0051]

[0052] , where the offset factor is the screening degree value used to promote or slow down the muscle coordination data in the neural network model, the first positive coordination data: represents the data that plays a positive role in muscle synergy, the first reverse coordination data: represents the data that plays a negative role in muscle synergy; the formula calculates the ratio of the difference between the first positive coordination data and the first reverse coordination data to the sum of the two, and its value range is between -1 and 1; a positive value indicates that the positive coordination is dominant, a negative value indicates that the reverse coordination is dominant, and a value close to 0 indicates that the two coordinations roughly offset each other; where the positive coordination data in the formula also includes calculating the proportional relationship between the first positive coordination data and the first reverse coordination data, that is, whether the two satisfy a certain matching relationship,

[0053] First forward matching data = k × first reverse matching data + q

[0054] , where k is the proportional factor and q is the parameter. The formula calculates the proportional relationship between the first positive matching data and the first negative matching data. When the first positive matching data is proportional to the first negative matching data, the original value of the first positive matching data of the offset factor is changed to change the range of the offset factor. The specific calculation method of the first positive matching data in the offset factor is:

[0055] The first positive coordination data = Z × the first reverse coordination data, wherein the first reverse coordination data is normalized to -1 < first reverse coordination data < 0, the role of Z is to ensure that the absolute value of the first positive coordination data is greater than the absolute value of the first reverse coordination data, Z is a minimum value greater than the integer 1, so the range of the offset factor is from negative infinity to 0; wherein, in the neural network model, firstly, the input data of the muscle coordination data is recorded, and at the same time, the muscle coordination data passing through the offset factor is recorded, and then the fuzzy reasoning model is constructed, and finally the rehabilitation training evaluation of the rehabilitation personnel is carried out, and the formula is

[0056]

[0057] , where P is the evaluation score, q, q1, and q2 are three different weights, and the effective ratios of the coordinated muscle data and the muscle coordination data calculated by the fuzzy reasoning model through the offset factors.

[0058] Among them, according to the degree of recovery of the elbow joint of the rehabilitated personnel, the elbow joint muscle coordination data of the rehabilitated personnel are gradually extracted, from the coordination of muscles in pairs, threes and threes to the overall coordination data of the muscles at the entire elbow joint, and the elbow joint muscle coordination data that meets the elbow joint muscle coordination data but does not meet the relevant adjustment angle of the elbow joint is screened out according to the transformer model. At this time, a neural network model is constructed to screen out the muscle coordination data that depends on electrode stimulation from the elbow joint muscle coordination data that meets the relevant adjustment angle of the elbow joint. In this process, an offset factor is introduced as the degree value of the neural network model screening. The role of the offset factor is to measure the degree value of the muscle coordination data screened by the neural network model, because although The offset factor can be positive or negative, but the overall rehabilitation process of the rehabilitated personnel is gradually restored. Therefore, when the first positive coordination data is not proportional to the first reverse coordination data, the muscle coordination data can pass the screening of the neural network model, but the time or efficiency of passing is different. If the offset factor is always positive, the muscle coordination data can pass quickly, but if the offset factor has positive and negative values, the time or efficiency of the muscle coordination data passing will change. By measuring the degree of screening of each muscle coordination data in the neural network model, and then measuring whether the degree of dependence on electrode stimulation in the muscle coordination data reaches the point where the training plan needs to be readjusted or re-evaluated, and when the first positive coordination data is proportional to the first reverse coordination data When the relationship is proportional, that is, when the value of the k-fold first reverse matching data plus a certain parameter is equal to the value of the first forward matching data, it indicates that the first forward matching data and the first reverse matching data have a negative impact on the rehabilitation personnel at this time. This negative impact is caused by the rehabilitation personnel's increasing dependence on electrode stimulation. Specifically, when the effectiveness of electrode stimulation causes the rehabilitation personnel to become dependent, this dependence aggravates the negative impact on the rehabilitation process, which in turn causes a negative impact on the rehabilitation plan, which in turn causes an imbalance in the rehabilitation plan, which again causes the rehabilitation personnel to stay at a certain stage for a long time. Among them, the imbalance of the rehabilitation plan and the rehabilitation personnel's long-term stay at a certain stage will cause the rehabilitation personnel to increase their dependence on electrode stimulation, which in turn causes the rehabilitation personnel to become more dependent on it. It depends on the effectiveness of electrode stimulation, thus forming a dependent cycle; in order to indicate that when the first positive coordination data is proportional to the first reverse coordination data, the muscle coordination data in the neural network model needs to be eliminated, so the offset factor is adjusted to the range of negative infinity to 0 by adjusting the numerical value of the first positive coordination data. When the first reverse coordination data is close to (-1), the offset factor tends to negative infinity, and when the first reverse coordination data is close to (0), the offset factor tends to (0), indicating that these data showing good muscle coordination do not meet the requirements. Finally, the rehabilitation training of the rehabilitation personnel is evaluated by the P value, and the muscle coordination data passed in the neural network model is combined with the expert experience in the fuzzy reasoning model to quantify it and then evaluate it.

[0059] Refine the rehabilitation personnel assessment report, and adjust the rehabilitation training plan according to the rehabilitation personnel assessment report, including: extract elbow rehabilitation training data according to the rehabilitation personnel assessment report, and collect lateral limb muscle detection data according to the corresponding relationship; select a dynamic adjustment lateral limb recovery plan according to the rehabilitation personnel's rehabilitation plan; obtain the offset factor data, and when the offset factor data is positive and the first positive coordination data is not proportional to the first negative coordination data, construct a double limb balance model and an RNN model; input the muscle coordination data training set of the injured elbow limb and the uninjured elbow limb into the double limb balance model, dynamically adjust the training plan of the injured elbow limb according to the output result of the double limb balance model, and conduct elbow training evaluation, input the rehabilitation personnel's rehabilitation process data training set into the RNN model, obtain the predicted recovery degree value, and perform normalization as the weight coefficient; the formulas used in the double limb balance model include,

[0060]

[0061] , where k1 and k3 are set according to the rehabilitation goals or physical characteristics of the rehabilitation personnel themselves, and k2 is the weight coefficient obtained by normalizing the recovery degree value predicted by the RNN model; in actual applications, the values ​​of k1 and k3 are dynamically adjusted according to the rehabilitation goals or physical characteristics of the rehabilitation personnel themselves.

[0062] Among them, k1 and k3 solve the different recovery goals and physical characteristics of different groups of people. When evaluating, the values ​​of k1 and k3 can be flexibly adjusted according to the recovery goals of the rehabilitation group (athletes need to recover quickly) or their own physical conditions (for example, left-handed people rely more on their left hand in life). In addition, the recovery degree of the rehabilitation personnel is quantified by the offset factor. The arrow symbol Used to indicate the balanced relationship between the left and right sides.

[0063] According to the adjusted rehabilitation training plan, the degree of recovery of the elbow joint of the rehabilitation personnel shall be evaluated again, including: implementing the adjusted rehabilitation training plan to ensure that each link of the rehabilitation training plan is effectively implemented; regularly evaluating the recovery of the elbow joint of the rehabilitation personnel, comprehensively evaluating the rehabilitation effect through objective indicators and subjective feelings, ensuring the effectiveness of the rehabilitation training plan, and making necessary adjustments based on the evaluation results.

[0064] According to the degree of recovery of the elbow joint of the rehabilitation personnel, a final evaluation report is obtained, including: according to the actual degree of recovery of the elbow joint of the rehabilitation personnel, a comprehensive and detailed comprehensive evaluation of the rehabilitation effect is carried out. The report records in detail the rehabilitation progress, existing problems and targeted follow-up suggestions.

[0065] Embodiment 2: Based on embodiment 1, a rehabilitation training evaluation system for an elbow joint includes:

[0066] Data collection module: used to obtain data on elbow injury type, physical characteristics, recovery type, and electrode usage of rehabilitation personnel;

[0067] Evaluation and adjustment module: used for LSTM model construction, dependency model construction, muscle synergy model construction, secondary evaluation model construction, data analysis, plan adjustment, and dynamic regulation;

[0068] Assessment execution and report generation module: used for rehabilitation training execution, assessment, comprehensive assessment, and report generation.

[0069] The above is a detailed introduction to the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea. At the same time, for general technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for evaluating rehabilitation training of an elbow joint, characterized in that: The following steps are involved: S1: Obtain the rehabilitation personnel's examination information and conduct a preliminary assessment of the rehabilitation personnel's elbow joint rehabilitation training; S2: Obtain the electrode usage of rehabilitation personnel and conduct secondary evaluation based on the electrode detection data; S3: Refine the rehabilitation personnel assessment report and adjust the rehabilitation training plan according to the rehabilitation personnel assessment report; S4: According to the adjusted rehabilitation training plan, re-evaluate the degree of recovery of the elbow joint of the rehabilitation personnel; S5: Obtain the final assessment report based on the rehabilitation personnel’s elbow recovery degree; The method of obtaining the electrode usage of the rehabilitation personnel and performing secondary evaluation based on the electrode detection data includes: collecting the electrode usage of the rehabilitation personnel; extracting the electrode stimulation data and the electrode detection data from the electrode usage data; constructing a dependency model using the electrode stimulation data; constructing a muscle synergy model using the electrode detection data; constructing a secondary evaluation model by combining the dependency model and the muscle synergy model; and obtaining a secondary evaluation report using the secondary evaluation model; The method of constructing a dependency model using electrode stimulation data includes: extracting electrode stimulation treatment effectiveness data and electrode stimulation treatment frequency data; calculating a first dependency of a rehabilitation person on electrode stimulation; evaluating the negative impact of the rehabilitation process and the imbalance of the rehabilitation plan; obtaining the length of stay in a certain stage of the rehabilitation phase; calculating a change in the degree of dependency of the rehabilitation person on electrode stimulation to obtain a second dependency; and constructing a dependency model by combining the first dependency and the second dependency. The method of combining the first dependency and the second dependency to construct a dependency model includes: extracting muscle synergy data in the muscle synergy model; obtaining muscle synergy data under the first dependency and muscle synergy data under the second dependency, defining the muscle synergy data under the first dependency as first positive coordination data, and defining the muscle synergy data under the second dependency as first negative coordination data; constructing the dependency model according to the first positive coordination data and the first negative coordination data; The method of constructing a dependency model based on the first forward coordination data and the first reverse coordination data includes: extracting muscle coordination data in the muscle coordination model, including coordination of muscles in pairs, in pairs, and up to the overall coordination data of muscles at the entire elbow joint; constructing a transformer model according to the rehabilitation stage of the rehabilitation personnel, and inputting the muscle coordination data training set into the transformer model for training; obtaining the adjustment angle of the elbow joint under muscle coordination; extracting muscle coordination data that meets the elbow joint adjustment angle of the rehabilitation personnel in the current rehabilitation stage; constructing a neural network model, inputting the muscle coordination data that meets the elbow joint adjustment angle of the rehabilitation personnel in the current rehabilitation stage, calculating the first forward coordination data and the first reverse coordination data to obtain the offset factor, and obtaining the muscle coordination data that can be offset by the offset factor within the stage time; analyzing the muscle coordination data that passes the offset factor to obtain an evaluation report of the rehabilitation personnel; The specific calculation formula of the offset factor is: , where the offset factor is the screening degree value used to promote or slow down the muscle coordination data in the neural network model, the first positive coordination data: represents the data that plays a positive role in muscle synergy, the first reverse coordination data: represents the data that plays a negative role in muscle synergy; the formula calculates the ratio of the difference between the first positive coordination data and the first reverse coordination data to the sum of the two, and its value range is between -1 and 1; a positive value indicates that the positive coordination is dominant, a negative value indicates that the reverse coordination is dominant, and a value close to 0 indicates that the two coordinations roughly offset each other; where the positive coordination data in the formula also includes calculating the proportional relationship between the first positive coordination data and the first reverse coordination data, that is, whether the two satisfy a certain matching relationship, ,in, is the scale factor, is a parameter, and the formula calculates the proportional relationship between the first positive matching data and the first negative matching data. When the first positive matching data is proportional to the first negative matching data, the original value of the first positive matching data of the offset factor is changed to change the range of the offset factor. The specific calculation method of the first positive matching data in the offset factor is: , where the first reverse matching data is normalized to -1<first reverse matching data<0, The role of ensures that the absolute value of the first forward matching data is greater than the absolute value of the first reverse matching data. is a minimum value greater than the integer 1, so the range of the offset factor is from negative infinity to 0; in the neural network model, firstly, the input data of the muscle coordination data is recorded, and at the same time, the muscle coordination data passing through the offset factor is recorded, then the fuzzy reasoning model is constructed, and finally the rehabilitation training evaluation of the rehabilitation personnel is carried out. The formula is , where P is the evaluation score value, , , The effective ratios of the coordinated muscle data and the muscle coordinated data through the offset factors are calculated by the fuzzy inference model for three different weights.

2. The method for evaluating rehabilitation training of an elbow joint according to claim 1, characterized in that: The obtaining of examination information of rehabilitation personnel and conducting preliminary evaluation of rehabilitation training for the rehabilitation personnel's elbow joint include: collecting rehabilitation personnel's elbow joint injury type data, physical characteristic data, and recovery type data, constructing an LSTM model, inputting the collected data into the LSTM model for training, and using the trained LSTM model to obtain preliminary evaluation results of rehabilitation training for the rehabilitation personnel's elbow joint.

3. The method for evaluating rehabilitation training of an elbow joint according to claim 1, characterized in that: The method of refining the rehabilitation personnel assessment report and adjusting the rehabilitation training plan according to the rehabilitation personnel assessment report includes: extracting elbow joint rehabilitation training data according to the rehabilitation personnel assessment report, and collecting lateral limb muscle detection data according to the corresponding relationship; selecting a dynamic adjustment lateral limb recovery plan according to the rehabilitation personnel's rehabilitation plan; obtaining offset factor data, and constructing a double limb balance model and an RNN model when the offset factor data is a positive value and the first positive coordination data is not proportional to the first negative coordination data; inputting the muscle coordination data training set of the injured elbow joint limb and the uninjured elbow joint limb into the double limb balance model, dynamically adjusting the training plan of the injured elbow joint limb according to the output result of the double limb balance model, and conducting elbow joint training evaluation, inputting the rehabilitation personnel's rehabilitation process data training set into the RNN model, obtaining a predicted recovery degree value, and performing normalization processing as a weight coefficient; the formula used in the double limb balance model includes ,in, , According to the rehabilitation goals or physical characteristics of the rehabilitation personnel, The weight coefficient is obtained by normalizing the recovery degree value predicted by the RNN model; in practical applications, it is dynamically adjusted according to the rehabilitation personnel's own rehabilitation goals or their own physical characteristics. , The numerical value of .

4. The method for evaluating rehabilitation training of an elbow joint according to claim 1, characterized in that: The described method of re-evaluating the degree of recovery of the elbow joint of the rehabilitated person according to the adjusted rehabilitation training plan includes: executing the adjusted rehabilitation training plan to ensure that each link of the rehabilitation training plan is effectively implemented; regularly evaluating the recovery of the elbow joint of the rehabilitated person, comprehensively evaluating the rehabilitation effect through objective indicators and subjective feelings, ensuring the effectiveness of the rehabilitation training plan, and making necessary adjustments based on the evaluation results.

5. The method for evaluating rehabilitation training of an elbow joint according to claim 1, characterized in that: The final evaluation report is obtained according to the degree of recovery of the elbow joint of the rehabilitation personnel, including: conducting a comprehensive and detailed comprehensive evaluation of the rehabilitation effect according to the actual degree of recovery of the elbow joint of the rehabilitation personnel, and the report records in detail the rehabilitation progress, existing problems and targeted follow-up suggestion information.

6. A rehabilitation training evaluation system for an elbow joint, according to a rehabilitation training evaluation method for an elbow joint according to any one of claims 1 to 5, comprising: Data collection module: used to obtain data on elbow injury type, physical characteristics, recovery type, and electrode usage of rehabilitation personnel; Evaluation and adjustment module: used for LSTM model construction, dependency model construction, muscle synergy model construction, secondary evaluation model construction, data analysis, plan adjustment, and dynamic regulation; Assessment execution and report generation module: used for rehabilitation training execution, assessment, comprehensive assessment, and report generation.

Images