A pelvic floor muscle rehabilitation training adjustment method and system based on data analysis

Abstract

The present application relates to the field of data processing technology, and in particular to a pelvic floor muscle rehabilitation training adjustment method and system based on data analysis, the method comprising: clustering the rehabilitation data of all users according to muscle strength status to obtain the status cluster to which the current user belongs; constructing a sensitivity matrix based on the rehabilitation data of the status cluster; determining the expected increment of each state indicator in the expected target according to the muscle strength status and standard state of the current user, and calculating the priority of each control parameter according to the sensitivity matrix and the expected target; adjusting each control parameter in the initial control instruction in descending order of priority until the adjusted initial control instruction can make the predicted increment of each state indicator reach the expected increment. Through the technical solution of the present application, the training effect of each control parameter on pelvic floor rehabilitation training can be comprehensively considered to achieve precise adjustment of pelvic floor muscle rehabilitation training.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a pelvic floor muscle rehabilitation training adjustment method and system based on data analysis. Background Art

[0002] Pelvic floor muscle rehabilitation currently relies mainly on the patient's own pelvic floor muscle contraction exercises. As an auxiliary device, the pelvic floor rehabilitation therapy device applies electrical stimulation to the pelvic floor muscles to increase neuromuscular excitability, stimulate pelvic floor muscle contraction, and thus achieve the purpose of pelvic floor muscle rehabilitation training.

[0003] At present, the patent application document with application publication number CN119274748A discloses a method and system for intelligent generation of home pelvic floor muscle rehabilitation training program, wherein the method includes: obtaining historical test data generated by the pelvic floor muscles of historical subjects during the whole training stage; performing model training based on the historical test data to obtain a global program generation model and a local generation model set; obtaining electromyographic test data generated by the pelvic floor muscles of the current subject during the whole training stage; and determining the target training stage that needs to be targetedly improved in the multiple local training stages according to the electromyographic test data; determining the local program generation model corresponding to the target training stage in the local generation model set, and determining the target training stage in the electromyographic test data. The method comprises the following steps: inputting the electromyographic test data into the global solution generation model for global prediction to obtain the global rehabilitation training solution for the current object; using the local solution generation model corresponding to the target training stage, performing local prediction based on the stage test data and the local stage evaluation data of the target training stage to obtain a targeted training solution suitable for the target training stage; adjusting the global rehabilitation training solution based on the targeted training solution suitable for the target training stage to obtain a target rehabilitation training solution suitable for the pelvic floor muscles of the current object.

[0004] The above method determines a rehabilitation training plan for the pelvic floor muscles suitable for the current subject based on a global plan generation model and multiple local generation models. However, the training process of multiple generation models has extremely high requirements on server performance and is costly. In addition, the training plan in the above method does not involve the effects of different control parameters of the pelvic floor rehabilitation therapy instrument on rehabilitation training, and cannot achieve precise adjustment of the control parameters during pelvic floor muscle rehabilitation training. Summary of the Invention

[0005] In order to solve the technical problem of being unable to achieve precise adjustment of control parameters during pelvic floor muscle rehabilitation training, the present application provides a pelvic floor muscle rehabilitation training adjustment method and system based on data analysis, which can comprehensively analyze the training effects of various control parameters on pelvic floor rehabilitation training and achieve precise adjustment of pelvic floor muscle rehabilitation training.

[0006] In a first aspect, the present application provides a pelvic floor muscle rehabilitation training adjustment method based on data analysis, the adjustment method comprising: clustering the rehabilitation data of all users according to muscle strength status, obtaining the status cluster to which the current user belongs, the rehabilitation data comprising muscle strength status, control instructions and muscle strength status increment, the control instructions comprising multiple control parameters, and the muscle strength status comprising multiple status indicators; constructing a sensitivity matrix based on the rehabilitation data of the status cluster, the sensitivity matrix comprising the sensitivity of each control parameter in the control instruction to each status indicator in the muscle strength status; determining an expected target based on the muscle strength status and standard status of the current user, calculating the priority of each control parameter based on the sensitivity matrix and the expected target, the expected target comprising the expected increment of each status indicator; adjusting each control parameter in the initial control instruction in descending order of priority until the adjusted initial control instruction can make the predicted increment of each status indicator reach the expected increment, the predicted increment of each status indicator being obtained by the regression model of each status indicator.

[0007] The rehabilitation data of all users are clustered according to their muscle strength status, and the status cluster to which the current user belongs is determined according to the muscle strength status of the current user; a sensitivity matrix is ​​constructed according to the rehabilitation data in the status cluster to reflect the sensitivity of each control parameter in the control instruction to each state indicator in the muscle strength status under the muscle strength status of the current user; further, the priority of each control parameter is calculated according to the expected goal of the rehabilitation training and the sensitivity matrix, and the priority can reflect the effectiveness of the rehabilitation training in achieving the expected goal after adjusting the control parameter; the control parameter with the largest priority is adjusted first to enable the effect of the rehabilitation training to quickly reach the expected goal, and the control parameters are adjusted in turn according to the priority until the rehabilitation training reaches the expected goal; in this way, the control parameters of the pelvic floor muscle rehabilitation therapy instrument are determined quickly and efficiently, and the precise adjustment of the pelvic floor muscle rehabilitation training is achieved, so that the pelvic floor muscle rehabilitation training can achieve the expected goal.

[0008] Preferably, obtaining the state cluster to which the current user belongs includes: taking the Euclidean distance between muscle strength states as the clustering distance, clustering the rehabilitation data of all users, and using the elbow method to determine the number of clusters to obtain multiple state clusters; calculating the similarity between the muscle strength state of the current user and the cluster center of each state cluster, and taking the state cluster with the largest similarity as the state cluster to which the current user belongs.

[0009] Rehabilitation data belonging to the same state cluster as the current user's muscle strength state is obtained from the rehabilitation data of all users to provide a data basis for the adjustment of subsequent pelvic floor muscle rehabilitation training and ensure the accuracy of rehabilitation training adjustment under the current user's muscle strength state.

[0010] Preferably, the control parameters are current intensity, stimulation frequency and pulse width.

[0011] Preferably, constructing a sensitivity matrix includes: constructing a regression model of any state indicator, the regression model including a bias and weights of each control parameter, the input of the regression model is the standardized control parameter, and the output is the increment of the state indicator; the normalized weight of the control parameter is used as the sensitivity of the control parameter to the state indicator.

[0012] The sensitivity matrix can intuitively reflect the training effect of each control parameter on each state indicator in the muscle strength state under the current muscle strength state of the user.

[0013] Preferably, the status indicator The regression model satisfies the relationship:

[0014] ;in, To control the number of parameters, Status indicator The control parameters in the regression model The weight of is the bias, Control parameters after standardization The value of Status indicator increment.

[0015] Preferably, the status indicator The method for constructing the regression model includes: for any rehabilitation data of the state cluster, taking the control parameters after normalization in the control instruction as input, and taking the state index in the muscle strength state increment as input. The increment of is used as the label to obtain sample data, and the least squares method is used to fit the sample data to determine the weights and bias in the regression model.

[0016] Preferably, the control parameter Priority for:

[0017] ; is the number of state indicators, is the sensitivity matrix Rank The value of the column, is the sensitivity matrix The sum of the absolute values ​​of the rows, Status indicators for expected goals expected increase.

[0018] The priority of each control parameter is calculated according to the expected goal, and the priority can reflect the effectiveness of the rehabilitation training in achieving the expected goal after adjusting the control parameters.

[0019] Preferably, each control parameter in the initial control instruction is adjusted in order of priority from large to small until the adjusted initial control instruction can make the predicted increment of each state indicator reach the expected increment, including: taking the control parameter with the highest priority as the target parameter; inputting any numerical value within the safety range of the target parameter and other control parameters in the initial control instruction into the regression model of each state indicator to obtain the predicted increment of each state indicator; taking the sum of the absolute values ​​of the differences between the predicted increment of each state indicator and the expected increment in the expected target as the objective function, and adjusting the target parameter to the value corresponding to the minimum value of the objective function; adjusting each control parameter in order of priority from large to small until the value of the objective function is less than a preset threshold, thereby obtaining the adjusted initial control instruction.

[0020] The greater the priority of a control parameter, the more likely it is that the rehabilitation training will achieve the desired goal after adjusting the control parameter. In order to quickly and accurately obtain the control parameters in rehabilitation training, the control parameters with higher priorities are adjusted first, saving time in determining the rehabilitation training control parameters.

[0021] Preferably, the method for obtaining the initial control instruction is: clustering the rehabilitation data in the state cluster to which the current user belongs according to the muscle strength state increment to obtain multiple increment clusters; and taking the average value of each control instruction in the increment cluster to which the expected target belongs as the initial control instruction.

[0022] The average value of each control instruction in the incremental cluster to which the expected target belongs is used as the initial control instruction. The initial control instruction can make the rehabilitation training close to the expected target. Adjustments are made based on the initial control instruction to obtain the time of the adjusted initial control instruction, and the adjusted initial control instruction can be obtained efficiently and quickly.

[0023] In the second aspect of the present application, a pelvic floor muscle rehabilitation training and adjustment system based on data analysis is also provided, including a processor and a memory, wherein the memory stores computer program instructions. When the computer program instructions are executed by the processor, a pelvic floor muscle rehabilitation training and adjustment method based on data analysis according to the first aspect of the present application is implemented.

[0024] The technical solution of this application has the following beneficial technical effects:

[0025] Cluster the rehabilitation data of all users according to their muscle strength status, and determine the state cluster to which the current user belongs based on their muscle strength status. Construct a sensitivity matrix based on the rehabilitation data in the state cluster to reflect the sensitivity of each control parameter in the control instruction to each state indicator in the muscle strength status under the current user's muscle strength status. For example, the control parameter Status indicators When the sensitivity is greater than 0, it means that by increasing the control parameter Can increase status indicators , and the larger the sensitivity value, the greater the sensitivity is. To increase the status indicator The better the effect; further, the priority of each control parameter is calculated according to the expected goal of the rehabilitation training and the sensitivity matrix. The priority can reflect the effectiveness of the rehabilitation training in achieving the expected goal after adjusting the control parameters; the control parameter with the largest priority is adjusted first, so that the effect of the rehabilitation training can quickly reach the expected goal, and the control parameters are adjusted in sequence according to the priority until the rehabilitation training reaches the expected goal; in this way, the control parameters of the pelvic floor muscle rehabilitation therapy instrument are determined quickly and efficiently, and the precise adjustment of the pelvic floor muscle rehabilitation training is achieved, so that the pelvic floor muscle rehabilitation training can achieve the expected goal. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a flow chart of a pelvic floor muscle rehabilitation training adjustment method based on data analysis according to an embodiment of the present application.

[0027] Figure 2 Schematic diagram of a pelvic floor muscle assessment report according to an embodiment of the present application.

[0028] Figure 3 This is a structural block diagram of a pelvic floor muscle rehabilitation training adjustment system based on data analysis according to an embodiment of the present application. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the described embodiments are part of the embodiments of this application. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without making creative efforts are within the scope of protection of this application.

[0030] According to the first aspect of the present application, the present application provides a pelvic floor muscle rehabilitation training adjustment method based on data analysis. Figure 1 This is a flow chart of a pelvic floor muscle rehabilitation training adjustment method based on data analysis according to an embodiment of the present application. Figure 1As shown, the pelvic floor muscle rehabilitation training adjustment method based on data analysis includes steps S101 to S104, which are described in detail below.

[0031] S101, clustering the rehabilitation data of all users according to muscle strength status to obtain the status cluster to which the current user belongs, wherein the rehabilitation data includes muscle strength status, control instructions, and muscle strength status increments, wherein the control instructions include multiple control parameters, and the muscle strength status includes multiple status indicators.

[0032] In one embodiment, before using the pelvic floor rehabilitation therapy device to perform rehabilitation training on the current user, the muscle strength state of the current user can be obtained by monitoring the pelvic floor electromyographic signal. The muscle strength state is used to represent the health state of the pelvic floor muscles. The muscle strength state includes multiple state indicators, and the state indicators are the index values ​​of the pre-resting stage, the rapid contraction stage, the tense contraction stage, the endurance contraction stage and the post-resting stage; wherein the state indicators are the index values ​​in the pelvic floor muscle assessment report, which are well-known technologies for those skilled in the art and will not be described in detail here. Figure 2 As shown, it is a schematic diagram of the pelvic floor muscle assessment report according to an embodiment of the present application. For example, the index values ​​of the pre-resting stage include the mean value and the coefficient of variation, and the index values ​​of the rapid contraction stage include the maximum value and the relaxation time. That is to say, the pelvic floor muscle assessment report includes a total of 11 index values.

[0033] Obtaining the state cluster to which the current user belongs includes: using the Euclidean distance between muscle strength states as the clustering distance, clustering the rehabilitation data of all users, and using the elbow method to determine the number of clusters to obtain multiple state clusters; calculating the similarity between the muscle strength state of the current user and the cluster center of each state cluster, and taking the state cluster with the largest similarity as the state cluster to which the current user belongs.

[0034] Among them, the rehabilitation data is the muscle strength state, control instructions and muscle strength state increment during the historical working process of the pelvic floor rehabilitation therapy instrument. The control instructions include multiple control parameters, which are current intensity, stimulation frequency and pulse width. The current intensity is used to control the stimulation intensity, which directly affects the degree of muscle fiber activation; the stimulation frequency is used to control the time interval between two stimulations, which indirectly affects the rhythmicity and contraction frequency of muscle reactions; the pulse width is used to adjust the stimulation depth, thereby affecting the penetration depth of electrical stimulation and the intensity of action on deep muscles.

[0035] In this way, before using the pelvic floor rehabilitation therapy apparatus to perform rehabilitation training on the current user, the muscle strength state of the current user is obtained, and rehabilitation data belonging to the same state cluster as the muscle strength state of the current user is obtained from the rehabilitation data of all users.

[0036] S102: constructing a sensitivity matrix based on the rehabilitation data of the state cluster, wherein the sensitivity matrix includes the sensitivity of each control parameter in the control instruction to each state indicator in the muscle strength state.

[0037] In one embodiment, under the current muscle strength state of the user, the rehabilitation effects of various control parameters on various state indicators in the muscle strength state are different. In order to quickly achieve the expected goals in the subsequent rehabilitation training process, it is necessary to judge the sensitivity of various control parameters to various state indicators in the muscle strength state under the current muscle strength state of the user.

[0038] Specifically, constructing a sensitivity matrix includes: constructing a regression model of any state indicator, the regression model includes a bias and weights of each control parameter, the input of the regression model is the standardized control parameter, and the output is the increment of the state indicator; the normalized weight of the control parameter is used as the sensitivity of the control parameter to the state indicator, thereby obtaining a sensitivity matrix.

[0039] Among them, the sensitivity is a value from -1 to 1. When the control parameter Status indicators When the sensitivity is equal to 0, it means adjusting the control parameters Cannot affect status indicators ; When the control parameters Status indicators When the sensitivity is greater than 0, it means that by increasing the control parameter Ability to increase status indicators , and the larger the sensitivity value, the greater the sensitivity is. To increase the status indicator The better the effect; when the control parameters Status indicators When the sensitivity is less than 0, it means that by increasing the control parameter Can reduce status indicators , and the larger the sensitivity value, the greater the sensitivity is. To reduce the status indicator The better the effect.

[0040] Among them, one state indicator corresponds to one regression model, which is used to predict the increment of the corresponding state indicator based on each control parameter. The regression model satisfies the relationship:

[0041] ;in, To control the number of parameters, Status indicator The control parameters in the regression model The weight of is the bias, Control parameters after standardization The value of Status indicator increment.

[0042] In the rehabilitation data of the state cluster, the least square method is used to fit the regression model of each state indicator to determine the weight and bias of each control parameter in the regression model. The method for constructing the regression model includes: for any rehabilitation data of the state cluster, taking the control parameters after normalization in the control instruction as input, and taking the state index in the muscle strength state increment as input. The increment of is used as the label to obtain sample data, and the least squares method is used to fit the sample data to determine the weights and bias in the regression model.

[0043] In other embodiments, the regression model may also adopt a fully connected neural network.

[0044] In this way, the construction of the sensitivity matrix is ​​completed. The sensitivity matrix has N rows and M columns, where N is the number of control parameters and M is the number of state indicators. Rank Column value For control parameters Status indicators The sensitivity matrix can intuitively reflect the training effect of each control parameter on each state indicator in the muscle strength state under the current user's muscle strength state.

[0045] S103, determining an expected target based on the current user's muscle strength state and standard state, and calculating the priority of each control parameter based on the sensitivity matrix and the expected target. The expected target includes the expected increment of each state indicator.

[0046] In one embodiment, the standard state is a reference value of each state indicator under normal circumstances. By comparing the current user's muscle strength state with the standard state, the expected increment of each state indicator can be obtained. That is, the expected target includes the expected increment of each state indicator.

[0047] For example, taking the variation value of the endurance contraction stage as an example, the variation value of the endurance contraction stage in the muscle strength state is 0.25, while the reference value corresponding to the variation value of the endurance contraction stage in the standard state is less than 0.2. Therefore, it can be determined that the expected target includes an expected increment of the variation value of the endurance contraction stage of -0.05.

[0048] In one embodiment, the priority of any control parameter is used to reflect the effectiveness of the rehabilitation training in achieving the expected goal after adjusting the corresponding control parameter. Priority for:

[0049] ; is the number of state indicators, is the sensitivity matrix Rank The value of the column, is the sensitivity matrix The sum of the absolute values ​​of the rows, Status indicators for expected goals expected increase.

[0050] Among them, when characterizing the relative size of the sensitivity of a control parameter to a state indicator, the sum of the absolute values ​​of the values ​​in the rows where the control parameters are located is different, which will lead to errors. Therefore, in the control parameters Before the priority, calculate the control parameters Status indicators Relative sensitivity , ensure that the sum of the relative sizes of the sensitivities of each state indicator in the row where each control parameter is located is 1, and accurately measure the control parameters priority.

[0051] Understandably, when the control parameters Status indicators Relative sensitivity is a positive number greater than 0, and the status indicator The expected increase in and status indicators The expected increment has the same sign), by adjusting the control parameters Enables status indicators The increment of is close to the expected increment, and The larger the value, the more control parameters should be adjusted. Able to effectively adjust status indicators , so that the status indicator Quickly approach the expected increment, at this time, adjust the control parameters Can quickly achieve the expected goal and control parameters Assign a higher priority; similarly, when the control parameter Status indicators Relative sensitivity is a positive number greater than 0, and the status indicator The expected increase in and status indicators When the expected increment has a different sign), by adjusting the control parameters Enable status indicators The increment is far away from the expected increment. At this time, adjust the control parameters Unable to achieve the expected goal, for control parameters Assign a priority less than 0.

[0052] In this way, the priority of each control parameter is calculated according to the expected goal, and the priority can reflect the effectiveness of the rehabilitation training in achieving the expected goal after adjusting the control parameter.

[0053] S104, adjusting each control parameter in the initial control instruction in descending order of priority until the adjusted initial control instruction can make the predicted increment of each state indicator reach the expected increment, and the predicted increment of each state indicator is obtained by the regression model of each state indicator.

[0054] In one embodiment, the greater the priority of a control parameter, the more likely it is that the rehabilitation training will achieve the expected goal after adjusting the control parameter; in order to quickly and accurately obtain the control parameters in rehabilitation training, the control parameters with higher priorities are adjusted first, saving time in determining the rehabilitation training control parameters.

[0055] Specifically, each control parameter in the initial control instruction is adjusted in order of priority from large to small until the adjusted initial control instruction can make the predicted increment of each state indicator reach the expected increment, including: taking the control parameter with the highest priority as the target parameter; inputting any numerical value within the safety range of the target parameter and other control parameters in the initial control instruction into the regression model of each state indicator to obtain the predicted increment of each state indicator; taking the sum of the absolute values ​​of the differences between the predicted increment of each state indicator and the expected increment in the expected target as the objective function, and adjusting the target parameter to the value corresponding to the minimum value of the objective function; adjusting each control parameter in order of priority from large to small until the value of the objective function is less than a preset threshold, thereby obtaining the adjusted initial control instruction.

[0056] The preset threshold is 0.5. In the state cluster of the current user, one state indicator corresponds to one regression model. Any value of the target parameter and other control parameters in the initial control instruction constitute a new initial control instruction. This new initial control instruction is input into the state indicator. The regression model can output the state index The predicted increment of each state indicator can be obtained similarly; when the sum of the absolute values ​​of the differences between the predicted increment of each state indicator and the expected increment in the expected target is less than the preset threshold, it means that the initial control instruction at this time can make the predicted increment of each state indicator reach the expected increment, and the adjusted initial control instruction is obtained.

[0057] Among them, the safety range of the target parameter is pre-set to avoid safety risks for users due to excessive values ​​of the control parameters; the initial control instructions are randomly generated.

[0058] In one embodiment, the method for obtaining the initial control instruction is: clustering the rehabilitation data in the state cluster to which the current user belongs according to the muscle strength state increment to obtain multiple increment clusters; and taking the average value of each control instruction in the increment cluster to which the expected target belongs as the initial control instruction.

[0059] The muscle strength state and the muscle strength state increment in the incremental cluster to which the expected target belongs are basically consistent; the average value of each control instruction in the incremental cluster to which the expected target belongs is used as the initial control instruction, and the initial control instruction can be close to the expected target. Adjustments are made based on the initial control instruction, and the time for obtaining the adjusted initial control instruction can be obtained, and the adjusted initial control instruction can be obtained efficiently and quickly.

[0060] In this way, the adjusted initial control instructions are used to control the pelvic floor muscle rehabilitation therapy instrument, thereby achieving precise adjustment of the pelvic floor muscle rehabilitation training, and quickly and efficiently determining the various control parameters of the pelvic floor muscle rehabilitation therapy instrument, thereby achieving precise adjustment of the pelvic floor muscle rehabilitation training.

[0061] According to the second aspect of the present application, the present application also provides a pelvic floor muscle rehabilitation training adjustment system based on data analysis. Figure 3 This is a structural block diagram of a pelvic floor muscle rehabilitation training and adjustment system based on data analysis according to an embodiment of the present application. Figure 3 As shown, the system 50 includes a processor and a memory, wherein the memory stores computer program instructions. When the computer program instructions are executed by the processor, the method for pelvic floor muscle rehabilitation training and adjustment based on data analysis according to the first aspect of the present application is implemented. The system also includes other components familiar to those skilled in the art, such as a communication bus and a communication interface. The configuration and functions of these components are well known in the art and are therefore not described in detail here.

[0062] It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present application, and these all fall within the scope of protection of the present application.

Claims

1. A pelvic floor muscle rehabilitation training and adjustment method based on data analysis, characterized in that: The adjustment method comprises: Clustering the rehabilitation data of all users based on muscle strength status to obtain the status cluster to which the current user belongs, wherein the rehabilitation data includes muscle strength status, control instructions, and muscle strength status increments. The control instructions include multiple control parameters, and the muscle strength status includes multiple status indicators. The control parameters are current intensity, stimulation frequency, and pulse width. Constructing a sensitivity matrix based on the rehabilitation data of the state cluster, wherein the sensitivity matrix includes the sensitivity of each control parameter in the control instruction to each state indicator in the muscle strength state; The expected target is determined based on the current user's muscle strength state and standard state, and the priority of each control parameter is calculated based on the sensitivity matrix and the expected target. The expected target includes the expected increment of each state indicator; the control parameter with the highest priority is used as the target parameter; any numerical value within the safety range of the target parameter and other control parameters in the initial control instruction are input into the regression model of each state indicator respectively to obtain the predicted increment of each state indicator; the sum of the absolute value of the difference between the predicted increment of each state indicator and the expected increment in the expected target is used as the objective function, and the target parameter is adjusted to the value corresponding to the minimum value of the objective function; each control parameter is adjusted in order from high to low priority until the value of the objective function is less than the preset threshold, and the adjusted initial control instruction is obtained, and the predicted increment of each state indicator is obtained by the regression model of each state indicator.

2. A pelvic floor muscle rehabilitation training and adjustment method based on data analysis according to claim 1, characterized in that: Obtaining the current user's status cluster includes: The Euclidean distance between muscle strength states is used as the clustering distance to cluster the rehabilitation data of all users, and the elbow method is used to determine the number of clusters to obtain multiple state clusters; Calculate the similarity between the current user's muscle strength state and the cluster center of each state cluster, and take the state cluster with the greatest similarity as the state cluster to which the current user belongs.

3. The pelvic floor muscle rehabilitation training and adjustment method based on data analysis according to claim 1, characterized in that: Constructing a sensitivity matrix involves: A regression model for any state indicator is constructed, wherein the regression model includes a bias and weights of each control parameter. The input of the regression model is the normalized control parameter, and the output is the increment of the state indicator. The normalized weight of the control parameter is used as the sensitivity of the control parameter to the state indicator.

4. A pelvic floor muscle rehabilitation training and adjustment method based on data analysis according to claim 3, characterized in that: Status indicators The regression model satisfies the relationship: ;in, To control the number of parameters, Status indicator The control parameters in the regression model The weight of is the bias, Control parameters after standardization The value of Status indicator increment.

5. The pelvic floor muscle rehabilitation training and adjustment method based on data analysis according to claim 3, characterized in that: Status indicators The methods for constructing the regression model include: For any rehabilitation data of the state cluster, the standardized control parameters in the control instruction are used as input, and the state index in the muscle strength state increment is used as input. The increment of is used as the label to obtain sample data, and the least squares method is used to fit the sample data to determine the weights and bias in the regression model.

6. The pelvic floor muscle rehabilitation training and adjustment method based on data analysis according to claim 1, characterized in that: Control parameters Priority for: ; is the number of state indicators, is the sensitivity matrix Rank The value of the column, is the sensitivity matrix The sum of the absolute values ​​of the rows, Status indicators for expected goals expected increase.

7. The pelvic floor muscle rehabilitation training and adjustment method based on data analysis according to claim 1, characterized in that: The method for obtaining the initial control instruction is: clustering the rehabilitation data in the state cluster to which the current user belongs according to the muscle strength state increment to obtain multiple increment clusters; and taking the average value of each control instruction in the increment cluster to which the expected target belongs as the initial control instruction.

8. A pelvic floor muscle rehabilitation training and adjustment system based on data analysis, characterized in that: The method comprises a processor and a memory, wherein the memory stores computer program instructions, and when the computer program instructions are executed by the processor, a pelvic floor muscle rehabilitation training adjustment method based on data analysis according to any one of claims 1 to 7 is implemented.

Images