CN112957056A - Method and system for extracting muscle fatigue grade features by utilizing cooperative network - Google Patents
Method and system for extracting muscle fatigue grade features by utilizing cooperative network Download PDFInfo
- Publication number
- CN112957056A CN112957056A CN202110280463.8A CN202110280463A CN112957056A CN 112957056 A CN112957056 A CN 112957056A CN 202110280463 A CN202110280463 A CN 202110280463A CN 112957056 A CN112957056 A CN 112957056A
- Authority
- CN
- China
- Prior art keywords
- network
- fatigue
- different
- channel
- muscle fatigue
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/45—For evaluating or diagnosing the musculoskeletal system or teeth
- A61B5/4519—Muscles
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/725—Details of waveform analysis using specific filters therefor, e.g. Kalman or adaptive filters
Abstract
The invention relates to a method and a system for extracting muscle fatigue grade characteristics by utilizing a cooperative network, wherein the method comprises the following steps: dividing a plurality of muscles on a body into a plurality of channels, collecting myoelectric data corresponding to each channel, and preprocessing the collected myoelectric data; calculating the Pearson correlation coefficient among different channels, and constructing a cooperative network diagram among different channels; analyzing the difference among different fatigue grades and extracting features, wherein the features comprise channel relation feature extraction and network structure feature extraction; and combining the extracted channel relation characteristics and the network structure characteristics to jointly form a characteristic vector of the muscle fatigue grade. The invention can effectively solve the problem of poor robustness of the characteristics extracted by muscle fatigue, the network graph can visually display the characteristic difference among different fatigue grades, and the variance analysis and the network parameters can give quantitative results on numerical values, thereby having good reliability.
Description
Technical Field
The invention relates to the technical field of rehabilitation medicine and human ergonomics, in particular to a method and a system for extracting muscle fatigue grade characteristics by utilizing a cooperative network.
Background
For scenes needing long-time work, such as driving an automobile, loading and unloading goods and the like, muscles need to be in a contraction state for a long time, fatigue can occur in a continuous state, at the moment, if a person wants to maintain the original level output muscle strength or work, muscle damage can be caused, and muscle fatigue can be accumulated continuously to finally cause musculoskeletal damage. In order to avoid the injury caused by muscle fatigue as much as possible, in the fields of rehabilitation medicine, human ergonomics and the like, a corresponding rehabilitation plan or workload needs to be made by detecting the fatigue state of human muscles, so that the research on the feature extraction method of muscle fatigue has important application value in the fields of rehabilitation medicine, human ergonomics and the like.
Currently, the muscle fatigue state can be evaluated by extracting surface electromyographic signal features. Surface electromyography (sEMG) is a non-stationary microelectric signal whose signal characteristics include changes in muscle fatigue status, and generally uses the sEMG to evaluate the time-domain and frequency-domain characteristics of multiple signals used for muscle fatigue, wherein the time-domain characteristics include root mean square, mean and mean rectification values, etc.; the frequency domain features include median and mean frequencies. However, time-frequency features to assess muscle fatigue are less robust to the complex problem of cross-task multi-classification.
Disclosure of Invention
Therefore, the technical problem to be solved by the invention is to overcome the defect that the robustness of the time-frequency characteristic evaluation complex muscle fatigue problem in the prior art is poor, so that the muscle fatigue grade characteristic extraction method and the muscle fatigue grade characteristic extraction system which have good robustness and utilize the cooperative network are provided.
In order to solve the above technical problem, the method for extracting muscle fatigue level features by using a collaborative network according to the present invention comprises the following steps: step S1: dividing a plurality of muscles on a body into a plurality of channels, collecting myoelectric data corresponding to each channel, and preprocessing the collected myoelectric data; step S2: calculating Pearson correlation coefficients among different channels according to the preprocessed electromyographic data, and constructing a cooperative network diagram among different channels according to the Pearson correlation coefficients; step S3: analyzing differences among different fatigue grades through a collaborative network diagram and extracting characteristics, wherein when channel relation characteristics are extracted, Pearson correlation coefficients among different channels form a column matrix, single-factor unitary variance analysis is adopted to obtain variance analysis results among the channels, channel pairs with significant differences among the different fatigue grades are distinguished according to probability values P in the analysis results, and the correlation coefficients of the distinguished channel pairs are used as muscle fatigue grade characteristics; when extracting the network structure characteristics, calculating network parameters of a clustering coefficient, network efficiency and network density, and taking the network parameters as muscle fatigue grade characteristics; step S4: and combining the extracted channel relation characteristics and the network structure characteristics to jointly form a characteristic vector of the muscle fatigue grade.
In an embodiment of the invention, the collecting of the electromyographic data corresponding to each channel is performed by a cross-task back force meter.
In one embodiment of the invention, the cross-task refers to physical training and fatigue driving, surface electromyographic signal data of the two tasks are combined during data acquisition, the fatigue state of the subjective representation of the testee is determined through subjective scale questionnaire survey, and the fatigue state is divided into no fatigue, moderate fatigue and severe fatigue.
In an embodiment of the present invention, a method for preprocessing collected electromyographic data includes: and performing dangerous wave processing on the data by applying band-pass filtering, and then performing normalization processing to obtain electromyographic data for solving the Pearson correlation coefficient.
In one embodiment of the invention, when the pearson correlation coefficients between different channels are calculated, if the pearson correlation coefficient value is greater than 0, the positive correlation between the variation trends of the two channels is represented; if the Pearson correlation coefficient value is less than 0, the change trend of the two channels is in negative correlation; if the correlation coefficient approaches 0, it indicates that the two channels are uncorrelated.
In an embodiment of the present invention, the method for constructing the cooperative network diagram between different channels according to the pearson correlation coefficient includes: and (4) carrying out difference on the average values of the correlation coefficients of different fatigue grades, and constructing a network change diagram among different fatigue grades according to the difference values.
In one embodiment of the invention, when the one-factor univariate variance analysis is adopted, the Pearson correlation coefficients among different channels form a column matrix and correspond to the labels of fatigue levels divided in advance.
In one embodiment of the invention, if the probability value P is less than 0.05, the channel has significant difference between different fatigue levels, and if the probability value P is not less than 0.05, the channel has no significant difference between different fatigue levels.
In an embodiment of the present invention, when the network parameters of the clustering coefficient, the network efficiency, and the network density are calculated, the obvious differences of the three network parameters at different fatigue levels are compared, and the differences between different fatigue levels are determined by calculating the network parameters of the clustering coefficient, the network efficiency, and the network density under the condition that the correlation coefficient takes the threshold value.
The invention also provides a system for extracting muscle fatigue grade characteristics by utilizing the cooperative network, which comprises the following steps: the acquisition preprocessing module is used for dividing a plurality of muscles on the body into a plurality of channels, acquiring myoelectric data corresponding to each channel and preprocessing the acquired myoelectric data; the cooperative network construction module is used for calculating the Pearson correlation coefficient among different channels according to the preprocessed electromyographic data and constructing a cooperative network diagram among the different channels according to the Pearson correlation coefficient; the analysis and extraction module is used for analyzing the differences among different fatigue grades through the collaborative network diagram and extracting the characteristics, wherein when the channel relation characteristics are extracted, the Pearson correlation coefficients among different channels form a column matrix, the single-factor unitary variance analysis is adopted to obtain variance analysis results among the channels, channel pairs with significant differences among the different fatigue grades are distinguished according to the probability value P in the analysis results, and the correlation coefficients of the distinguished channel pairs are used as muscle fatigue grade characteristics; when extracting the network structure characteristics, calculating network parameters of a clustering coefficient, network efficiency and network density, and taking the network parameters as muscle fatigue grade characteristics; and the combination module is used for combining the extracted channel relation characteristics and the network structure characteristics to jointly form the characteristic vector of the muscle fatigue grade.
Compared with the prior art, the technical scheme of the invention has the following advantages:
the muscle fatigue grade characteristic extraction method and system utilizing the cooperative network provided by the invention construct the muscle cooperative network based on the correlation coefficient among different channels of electromyographic signals, judge the difference between fatigue grades through two methods of single-factor variance analysis and network parameter calculation, extract the channel relation characteristic and the network structure characteristic, and combine the channel relation characteristic and the network structure characteristic into the characteristic vector of the muscle fatigue grade. The method can effectively solve the problem of poor robustness of the muscle fatigue extraction features, the network diagram can visually show the feature difference among different fatigue grades, and the variance analysis and the network parameters can give quantitative results on numerical values, so that the method has good reliability.
Drawings
In order that the present disclosure may be more readily and clearly understood, reference is now made to the following detailed description of the embodiments of the present disclosure taken in conjunction with the accompanying drawings, in which
FIG. 1 is a flow chart of a muscle fatigue level feature extraction method using a collaborative network according to the present invention;
FIG. 2 is a schematic diagram of the fatigue-free cooperative network connection of the present invention;
FIG. 3 is a schematic diagram of a moderately fatigued cooperative network connection in accordance with the present invention;
FIG. 4 is a schematic diagram of the heavily fatigued collaborative network connection of the present invention;
FIG. 5 is a schematic diagram of the cooperative networking of the present invention between no fatigue and moderate fatigue;
FIG. 6 is a schematic diagram of the cooperative networking of the present invention between mild fatigue and severe fatigue;
FIG. 7 is a table of channel pair information with significant differences in the fatigue-free and moderate-fatigue single-factor univariate analysis of variance of the present invention;
FIG. 8 is a table of channel pair information for the single factor univariate analysis of severe fatigue and moderate fatigue of the present invention with significant differences;
FIG. 9 is a table comparing the clustering coefficients, network efficiency and network density for different fatigue levels according to the present invention.
Detailed Description
Example one
As shown in fig. 1, the present embodiment provides a method for extracting muscle fatigue level features using a collaborative network, and the step S1: dividing a plurality of muscles on a body into a plurality of channels, collecting myoelectric data corresponding to each channel, and preprocessing the collected myoelectric data; step S2: calculating Pearson correlation coefficients among different channels according to the preprocessed electromyographic data, and constructing a cooperative network diagram among different channels according to the Pearson correlation coefficients; step S3: analyzing differences among different fatigue grades through a collaborative network diagram and extracting characteristics, wherein when channel relation characteristics are extracted, a column matrix is formed by Pearson correlation coefficients among different channels, single-factor unitary variance analysis is carried out to obtain variance analysis results among the channels, channel pairs with significant differences among different fatigue grades are distinguished through probability values P in the analysis results, and the correlation coefficients of the distinguished channel pairs are used as muscle fatigue grade characteristics; when extracting the network structure characteristics, calculating network parameters of a clustering coefficient, network efficiency and network density, and taking the network parameters as muscle fatigue grade characteristics; step S4: and combining the extracted channel relation characteristics and the network structure characteristics to jointly form a characteristic vector of the muscle fatigue grade.
In the method for extracting muscle fatigue level features using a collaborative network according to this embodiment, in step S1, a plurality of muscles on a body are divided into a plurality of channels, myoelectric data corresponding to each channel are collected, and the collected myoelectric data are preprocessed, which is beneficial to removing interference; in step S2, the pearson correlation coefficients between different channels are calculated according to the preprocessed electromyographic data, constructing a cooperative network diagram among different channels according to the Pearson correlation coefficient, wherein the cooperative network is one of complex networks and is an angle and a method for researching complex problems, the method for researching the complex network is mainly based on graph theory, focuses on a topological structure of individual mutual association in a system, utilizes the correlation coefficient of the electromyographic signals to construct the cooperative network, the network nodes are corresponding muscles, the links between the nodes are correlation coefficients between the two muscles so as to represent the cooperative relationship between different muscles, the correlation coefficients of different fatigue states are different, so that the constructed cooperative network is also different, the muscle fatigue is evaluated, the different fatigue states can be effectively distinguished, the fatigue grade difference analysis method established by the correlation coefficient cooperative network has good stability; in the step S3, analyzing differences between different fatigue grades through a collaborative network diagram and extracting features, wherein when extracting channel relation features, a column matrix is formed by pearson correlation coefficients between different channels, single-factor unitary variance analysis is performed to obtain variance analysis results between the channels, channel pairs with significant differences between different fatigue grades are distinguished through probability values P in the analysis results, and the correlation coefficients of the distinguished channel pairs are used as muscle fatigue grade features; when extracting the network structure characteristics, calculating network parameters of a clustering coefficient, network efficiency and network density, and taking the network parameters as muscle fatigue grade characteristics to extract the muscle fatigue grade characteristics; in the step S4, the extracted channel relation features and network structure features are combined to form a feature vector of muscle fatigue level together, so that the problem of poor robustness of the muscle fatigue extracted features can be effectively solved, the network diagram can visually display the feature difference between different fatigue levels, the variance analysis and the network parameters can give quantitative results on the value, and the reliability is good.
In step S1, seven muscles are usually selected from the muscles of the body, including the upper trapezius, the middle trapezius, the lower trapezius, the erector spinae, the latissimus dorsi, the deltoid posterior cluster and the multifidus, corresponding to channels 1 to 7.
And acquiring the electromyographic data corresponding to each channel by a cross-task back force meter. Specifically, the cross-task refers to physical training and fatigue driving, surface electromyographic signal data of the two tasks are combined during data acquisition, the fatigue state of the subjective representation of the testee is determined through subjective scale questionnaire survey, and the fatigue state is divided into no fatigue, moderate fatigue and severe fatigue.
During specific collection, the electrode plate is attached to the measured muscle and connected with the myoelectricity collection device, the sampling frequency is 1024Hz, the sampling time is 7 seconds, questionnaires are conducted before and after the experiment, the questionnaires comprise subjective evaluation scales and behavioural tests, corresponding experiment data are divided into no fatigue (level 1), moderate fatigue (level 2) and severe fatigue (level 3) through the comprehensive questionnaire results.
The method for preprocessing the collected electromyographic data comprises the following steps: and performing dangerous wave processing on the data by applying band-pass filtering, and then performing normalization processing to obtain electromyographic data for solving the Pearson correlation coefficient. Specifically, Butterworth band-pass filtering is applied, the frequency of a pass band is set to be 10-350Hz, power frequency interference is removed through 50Hz dangerous wave processing, and then data are normalized to obtain electromyographic data for solving the Pearson correlation coefficient. Wherein the normalization adopts a formula: newValue ═ mean/std, where oldValue denotes the value before normalization of a certain data, newValue denotes the value after normalization of a certain data, mean denotes the mean of a data set, and std denotes the standard deviation of a data set.
In step S2, when calculating the pearson correlation coefficients between different channels according to the preprocessed electromyographic data, the formula of the pearson correlation coefficients is as follows:
when the Pearson correlation coefficients among different channels are calculated, if the Pearson correlation coefficient value is larger than 0, the positive correlation of the variation trends of the two channels is represented; if the Pearson correlation coefficient value is less than 0, the change trend of the two channels is in negative correlation; if the correlation coefficient approaches 0, it indicates that the two channels are uncorrelated.
The method for constructing the cooperative network diagram among different channels according to the Pearson correlation coefficient comprises the following steps: and (4) carrying out difference on the average values of the correlation coefficients of different fatigue grades, and constructing a network change diagram among different fatigue grades according to the difference values.
Constructing a cooperative network diagram among different channels of three fatigue grades through a Pearson correlation coefficient, wherein the specific method comprises the following steps: averaging the correlation coefficients of each fatigue grade, and drawing a network graph by using the average values; in order to further analyze the difference between different fatigue grades, a network change diagram between every two fatigue grades is also constructed, and the specific method comprises the following steps: and (4) carrying out difference on the average value of the correlation coefficients of the two grades, and drawing two difference value network graphs of moderate fatigue and no fatigue and severe fatigue and moderate fatigue by using the difference value.
Obtaining a correlation coefficient matrix of each tested person by calculating the pearson correlation coefficient, averaging the correlation coefficients of each grade, and drawing a collaborative network diagram by using the average value, as shown in the attached figures 2-4; the average values of the correlation coefficients of the two levels are subtracted, and a network diagram of the difference values between moderate fatigue and no fatigue is drawn by using the difference values and is shown in fig. 5, and a network diagram of the difference values between severe fatigue and moderate fatigue is shown in fig. 6.
In step S3, the pearson correlation coefficients between different channels are combined into a column matrix, and a one-factor unitary variance analysis is performed, so as to obtain a variance analysis result between each channel. In order to improve the reliability of the analysis result, when the single-factor univariate variance analysis is carried out, the Pearson correlation coefficients among different channels form a column matrix and correspond to the labels of fatigue grades divided in advance.
When the channel relation characteristics are extracted by adopting the one-factor variance analysis, the correlation coefficients of each channel pair of the three fatigue grades are arranged into a sequence and correspond to the corresponding grade serial numbers, the two columns of data are subjected to the one-factor one-element variance analysis, the channel pairs with significant differences among different fatigue grades are distinguished through the probability value P in the analysis result, and the correlation coefficients of the distinguished channel pairs are used as the channel relation characteristics of the muscle fatigue grades.
If the probability value P is less than 0.05, the channel has significant difference between different fatigue grades, and if the probability value P is not less than 0.05, the channel has no significant difference between different fatigue grades.
The invention distinguishes channel pairs with significant differences among different fatigue grades by analyzing probability values in results, and the information result of the channel pairs with significant differences in single-factor univariate variance analysis without fatigue and moderate fatigue is calculated by the method as shown in fig. 7, wherein the information result of the channel pairs with significant differences is shown in fig. 8, and the results of the channel pairs with significant differences are shown in fig. 1 and 2(P ═ 0.034), the channel pairs with significant differences in channel 1 and 4(P ═ 0.0013), the channel pairs with significant differences in channel 4 and 7(P ═ 0.0013) and the channel pairs with significant differences in channel 5 and 7(P ═ 1.0152), the channel pairs with significant differences in channel 1 and 2(P ═ 0.0004), the channel channels 1 and 4(P ═ 0.0092), the channel pairs with significant differences in channel 3 and 4(P ═ 0.0067), the channel pairs with significant differences in channel 5 and channel pair with significant differences in channel 7(P ═ 3.4 e-05).
When the network structure characteristics are extracted, the network parameters of the collaborative network are calculated, the network structure characteristics are extracted, the Pearson correlation coefficients of the same fatigue level are averaged, the collaborative network is made by using the obtained average correlation coefficient matrix, and the network parameter calculation is carried out, so that the network parameters are divided into a positive-direction weighted network and a negative-direction weighted network during the calculation due to the fact that the correlation coefficients have negative correlation.
The network parameters mainly comprise clustering coefficients, network efficiency and network density, and the corresponding calculation method comprises the following steps:
the clustering coefficient is:
wherein k isiIs the number of neighbors of node i, EiIs the number of edges actually existing between k neighbors of node i, and N is the number of nodes.
The network efficiency is as follows:
wherein d isijIs the distance between two nodes, and N is the number of nodes.
The network density is:
wherein, sigma VkIs the sum of all weighted edges and N is the number of nodes.
When the network parameters of the clustering coefficient, the network efficiency and the network density are calculated, the obvious differences of the three network parameters under different fatigue grades are compared, and the differences among different fatigue grades are judged by calculating the network parameters of the clustering coefficient, the network efficiency and the network density under the condition that the correlation coefficient takes the threshold value.
Specifically, network parameters of the collaborative network are calculated to extract network structure features, and the three network parameters, namely the clustering coefficient, the network efficiency and the network density, have obvious differences in different fatigue levels, and the clustering coefficient, the network efficiency and the network density are calculated to judge the differences among different fatigue levels under the condition that the correlation coefficient takes the threshold value of 0.05, and are used as the network structure features of muscle fatigue levels. And combining the extracted network structure characteristics and the channel relation characteristics to jointly form a characteristic vector of the muscle fatigue level.
The result of the network parameters obtained in the most general way is shown in fig. 9, and the variation and difference of the network parameters between different fatigue levels can be obtained from the result, for example, the network efficiency of the forward authorized network is from 0.271 to 0.267 to 0.271, the network efficiency is firstly decreased and then increased, the network efficiency is used as the network structure characteristic, and the extracted network structure characteristic and the channel relation characteristic are combined to form the characteristic vector of the muscle fatigue level together.
Example two
Based on the same inventive concept, this embodiment provides a system for extracting muscle fatigue level features using a collaborative network, the principle of solving the problem is similar to the method for extracting muscle fatigue level features using a collaborative network, and the repetition parts are not repeated, and the system specifically includes:
the acquisition preprocessing module is used for dividing a plurality of muscles on the body into a plurality of channels, acquiring myoelectric data corresponding to each channel and preprocessing the acquired myoelectric data;
the cooperative network construction module is used for calculating the Pearson correlation coefficient among different channels according to the preprocessed electromyographic data and constructing a cooperative network diagram among the different channels according to the Pearson correlation coefficient;
the analysis and extraction module is used for analyzing the differences among different fatigue grades through the collaborative network diagram and extracting the characteristics, wherein when the channel relation characteristics are extracted, the Pearson correlation coefficients among different channels form a column matrix, the single-factor unitary variance analysis is adopted to obtain variance analysis results among the channels, channel pairs with significant differences among the different fatigue grades are distinguished according to probability values in the analysis results, and the correlation coefficients of the distinguished channel pairs are used as muscle fatigue grade characteristics; when extracting the network structure characteristics, calculating network parameters of a clustering coefficient, network efficiency and network density, and taking the network parameters as muscle fatigue grade characteristics;
and the combination module is used for combining the extracted channel relation characteristics and the network structure characteristics to jointly form the characteristic vector of the muscle fatigue grade.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may 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, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may 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, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
It should be understood that the above examples are only for clarity of illustration and are not intended to limit the embodiments. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. And are neither required nor exhaustive of all embodiments. And obvious variations or modifications therefrom are within the scope of the invention.
Claims (10)
1. A muscle fatigue grade feature extraction method using a collaborative network is characterized by comprising the following steps:
step S1: dividing a plurality of muscles on a body into a plurality of channels, collecting myoelectric data corresponding to each channel, and preprocessing the collected myoelectric data;
step S2: calculating Pearson correlation coefficients among different channels according to the preprocessed electromyographic data, and constructing a cooperative network diagram among different channels according to the Pearson correlation coefficients;
step S3: analyzing differences among different fatigue grades through a collaborative network diagram and extracting characteristics, wherein when channel relation characteristics are extracted, Pearson correlation coefficients among different channels form a column matrix, single-factor unitary variance analysis is adopted to obtain variance analysis results among the channels, channel pairs with significant differences among different fatigue grades are distinguished according to probability values in the analysis results, and the correlation coefficients of the distinguished channel pairs are used as muscle fatigue grade characteristics; when extracting the network structure characteristics, calculating network parameters of a clustering coefficient, network efficiency and network density, and taking the network parameters as muscle fatigue grade characteristics;
step S4: and combining the extracted channel relation characteristics and the network structure characteristics to jointly form a characteristic vector of the muscle fatigue grade.
2. The method for extracting muscle fatigue level characteristics using a cooperative network according to claim 1, wherein: and acquiring the electromyographic data corresponding to each channel by a cross-task back force meter.
3. The method for extracting muscle fatigue level characteristics using a cooperative network according to claim 2, wherein: the cross-task refers to physical training and fatigue driving, surface electromyographic signal data of the two tasks are combined during data acquisition, the fatigue state of the subjective representation of the testee is determined through subjective scale questionnaire survey, and the fatigue state is divided into no fatigue, moderate fatigue and severe fatigue.
4. The method for extracting muscle fatigue level characteristics using a cooperative network according to claim 1, wherein: the method for preprocessing the collected electromyographic data comprises the following steps: and performing dangerous wave processing on the data by applying band-pass filtering, and then performing normalization processing to obtain electromyographic data for solving the Pearson correlation coefficient.
5. The method for extracting muscle fatigue level characteristics using a cooperative network according to claim 1, wherein: when the Pearson correlation coefficients among different channels are calculated, if the Pearson correlation coefficient value is larger than 0, the positive correlation of the variation trends of the two channels is represented; if the Pearson correlation coefficient value is less than 0, the change trend of the two channels is in negative correlation; if the correlation coefficient approaches 0, it indicates that the two channels are uncorrelated.
6. The method for extracting muscle fatigue level characteristics using a cooperative network according to claim 1 or 5, wherein: the method for constructing the cooperative network diagram among different channels according to the Pearson correlation coefficient comprises the following steps: and (4) carrying out difference on the average values of the correlation coefficients of different fatigue grades, and constructing a network change diagram among different fatigue grades according to the difference values.
7. The method for extracting muscle fatigue level characteristics using a cooperative network according to claim 1, wherein: and when the single-factor unitary variance analysis is adopted, the Pearson correlation coefficients among different channels form a column matrix and correspond to the labels of fatigue grades divided in advance.
8. The method for extracting muscle fatigue level characteristics using a cooperative network according to claim 1, wherein: if the probability value P is less than 0.05, the channel has significant difference between different fatigue grades, and if the probability value P is not less than 0.05, the channel has no significant difference between different fatigue grades.
9. The method for extracting muscle fatigue level characteristics using a cooperative network according to claim 1, wherein: when the network parameters of the clustering coefficient, the network efficiency and the network density are calculated, the obvious differences of the three network parameters under different fatigue grades are compared, and the differences among different fatigue grades are judged by calculating the network parameters of the clustering coefficient, the network efficiency and the network density under the condition that the correlation coefficient takes the threshold value.
10. A system for extracting muscle fatigue level characteristics using a cooperative network, comprising:
the acquisition preprocessing module is used for dividing a plurality of muscles on the body into a plurality of channels, acquiring myoelectric data corresponding to each channel and preprocessing the acquired myoelectric data;
the cooperative network construction module is used for calculating the Pearson correlation coefficient among different channels according to the preprocessed electromyographic data and constructing a cooperative network diagram among the different channels according to the Pearson correlation coefficient;
the analysis and extraction module is used for analyzing the differences among different fatigue grades through the collaborative network diagram and extracting the characteristics, wherein when the channel relation characteristics are extracted, the Pearson correlation coefficients among different channels form a column matrix, the single-factor unitary variance analysis is adopted to obtain variance analysis results among the channels, channel pairs with significant differences among the different fatigue grades are distinguished according to the probability value P in the analysis results, and the correlation coefficients of the distinguished channel pairs are used as muscle fatigue grade characteristics; when extracting the network structure characteristics, calculating network parameters of a clustering coefficient, network efficiency and network density, and taking the network parameters as muscle fatigue grade characteristics;
and the combination module is used for combining the extracted channel relation characteristics and the network structure characteristics to jointly form the characteristic vector of the muscle fatigue grade.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110280463.8A CN112957056B (en) | 2021-03-16 | 2021-03-16 | Method and system for extracting muscle fatigue grade features by utilizing cooperative network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110280463.8A CN112957056B (en) | 2021-03-16 | 2021-03-16 | Method and system for extracting muscle fatigue grade features by utilizing cooperative network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112957056A true CN112957056A (en) | 2021-06-15 |
| CN112957056B CN112957056B (en) | 2022-12-30 |
Family
ID=76277744
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110280463.8A Active CN112957056B (en) | 2021-03-16 | 2021-03-16 | Method and system for extracting muscle fatigue grade features by utilizing cooperative network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112957056B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113509151A (en) * | 2021-08-30 | 2021-10-19 | 北京理工大学 | Method and device for evaluating muscle tension level |
| CN114732424A (en) * | 2022-04-29 | 2022-07-12 | 苏州大学 | Method for extracting complex network attribute of muscle fatigue state based on surface electromyographic signal |
| CN114931390A (en) * | 2022-05-06 | 2022-08-23 | 电子科技大学 | Muscle force estimation method based on fatigue analysis |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002224072A (en) * | 2001-02-02 | 2002-08-13 | Mazda Motor Corp | Muscular fatigue evaluating device, muscular fatigue evaluating method, computer program for executing this method and storage medium for storing this computer program |
| CN105147251A (en) * | 2015-08-19 | 2015-12-16 | 宁波工程学院 | Muscle fatigue dynamic prediction method based on multi-channel sEMG |
| CN109662710A (en) * | 2018-12-06 | 2019-04-23 | 杭州电子科技大学 | A kind of EMG Feature Extraction based on convolutional neural networks |
| CN110522999A (en) * | 2019-09-29 | 2019-12-03 | 福州大学 | Electric stimulation with Fatigue Assessment function |
| CN111904416A (en) * | 2020-07-27 | 2020-11-10 | 深圳先进技术研究院 | Muscle fatigue detection method, computer device, and storage medium |
-
2021
- 2021-03-16 CN CN202110280463.8A patent/CN112957056B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002224072A (en) * | 2001-02-02 | 2002-08-13 | Mazda Motor Corp | Muscular fatigue evaluating device, muscular fatigue evaluating method, computer program for executing this method and storage medium for storing this computer program |
| CN105147251A (en) * | 2015-08-19 | 2015-12-16 | 宁波工程学院 | Muscle fatigue dynamic prediction method based on multi-channel sEMG |
| CN109662710A (en) * | 2018-12-06 | 2019-04-23 | 杭州电子科技大学 | A kind of EMG Feature Extraction based on convolutional neural networks |
| CN110522999A (en) * | 2019-09-29 | 2019-12-03 | 福州大学 | Electric stimulation with Fatigue Assessment function |
| CN111904416A (en) * | 2020-07-27 | 2020-11-10 | 深圳先进技术研究院 | Muscle fatigue detection method, computer device, and storage medium |
Non-Patent Citations (1)
| Title |
|---|
| 朱悦，李春光，郭浩，祝宇飞，孙立宁: "运动想象的机理研究与应用综述", 《计算机工程与应用》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113509151A (en) * | 2021-08-30 | 2021-10-19 | 北京理工大学 | Method and device for evaluating muscle tension level |
| CN113509151B (en) * | 2021-08-30 | 2022-03-25 | 北京理工大学 | Method and device for evaluating muscle tension level |
| CN114732424A (en) * | 2022-04-29 | 2022-07-12 | 苏州大学 | Method for extracting complex network attribute of muscle fatigue state based on surface electromyographic signal |
| CN114931390A (en) * | 2022-05-06 | 2022-08-23 | 电子科技大学 | Muscle force estimation method based on fatigue analysis |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112957056B (en) | 2022-12-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112957056B (en) | Method and system for extracting muscle fatigue grade features by utilizing cooperative network | |
| CN107095669B (en) | A kind of processing method and system of epileptic's EEG signals | |
| CN103034837B (en) | Characteristic parameter is associated with pulse condition key element | |
| CN110432898A (en) | A kind of epileptic attack eeg signal classification system based on Nonlinear Dynamical Characteristics | |
| CN101259016A (en) | Method for real time automatically detecting epileptic character wave | |
| CN102058413B (en) | Method for detecting EEG (electroencephalogram) alertness based on continuous wavelet transform | |
| CN104720796A (en) | Automatic detecting system and method for epileptic attack time period | |
| CN108959895B (en) | Electroencephalogram EEG (electroencephalogram) identity recognition method based on convolutional neural network | |
| CN105320969A (en) | A heart rate variability feature classification method based on multi-scale Renyi entropy | |
| CN109480833A (en) | The pretreatment and recognition methods of epileptic's EEG signals based on artificial intelligence | |
| CN103690160B (en) | Electroencephalogram feature extraction method based on non-Gaussian time sequence model | |
| Kołodziej et al. | A new method of EEG classification for BCI with feature extraction based on higher order statistics of wavelet components and selection with genetic algorithms | |
| CN105550659A (en) | Real-time electrocardiogramclassification method based on random projection | |
| CN107361765B (en) | Brain wave analysis method and device | |
| CN114732424B (en) | Method for extracting complex network attribute of muscle fatigue state based on surface electromyographic signal | |
| CN111000555B (en) | Training data generation method, automatic recognition model modeling method and automatic recognition method for epileptic electroencephalogram signals | |
| CN107595247A (en) | A kind of monitoring method and system of the depth of anesthesia based on EEG signals | |
| CN113868006A (en) | Time sequence detection method and device, electronic equipment and computer storage medium | |
| CN111248928A (en) | Pressure identification method and device | |
| CN112633195A (en) | Myocardial infarction identification and classification method based on frequency domain features and deep learning | |
| CN111067513B (en) | Sleep quality detection key brain area judgment method based on characteristic weight self-learning | |
| CN110032987B (en) | Surface electromyographic signal classification method based on cerebellar neural network model | |
| CN111956209A (en) | Electrocardiosignal R wave identification method based on EWT and structural feature extraction | |
| Yildirim et al. | Heartbeat type classification with optimized feature vectors | |
| CN116616709A (en) | Working memory load assessment method based on multichannel video in non-stationary state |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |