Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2022 Nov 29;13(12):2108.
doi: 10.3390/mi13122108.

A Novel Methodology for Classifying EMG Movements Based on SVM and Genetic Algorithms

Marcos Aviles  1 , Luz-María Sánchez-Reyes  1 , Rita Q Fuentes-Aguilar  2 , Diana C Toledo-Pérez  1 , Juvenal Rodríguez-Reséndiz  1
Affiliations

A Novel Methodology for Classifying EMG Movements Based on SVM and Genetic Algorithms

Marcos Aviles et al. Micromachines (Basel). .

Abstract

Electromyography (EMG) processing is a fundamental part of medical research. It offers the possibility of developing new devices and techniques for the diagnosis, treatment, care, and rehabilitation of patients, in most cases non-invasively. However, EMG signals are random, non-stationary, and non-linear, making their classification difficult. Due to this, it is of vital importance to define which factors are helpful for the classification process. In order to improve this process, it is possible to apply algorithms capable of identifying which features are most important in the categorization process. Algorithms based on metaheuristic methods have demonstrated an ability to search for suitable subsets of features for optimization problems. Therefore, this work proposes a methodology based on genetic algorithms for feature selection to find the parameter space that offers the slightest classification error in 250 ms signal segments. For classification, a support vector machine is used. For this work, two databases were used, the first corresponding to the right upper extremity and the second formed by movements of the right lower extremity. For both databases, a feature space reduction of over 65% was obtained, with a higher average classification efficiency of 91% for the best subset of parameters. In addition, particle swarm optimization (PSO) was applied based on right upper extremity data, obtaining an 88% average error and a 46% reduction for the best subset of parameters. Finally, a sensitivity analysis was applied to the characteristics selected by PSO and genetic algorithms for the database of the right upper extremity, obtaining that the parameters determined by the genetic algorithms show greater sensitivity for the classification process.

Keywords: electromyography; feature selection; metaheuristic algorithms; pattern recognition; support vector machine.

PubMed Disclaimer

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Figure 1
Figure 1
Electrical diagram of the complete EMG acquisition system. (a) The instrumentation amplifier is shown with the operational amplifier in integrating configuration, while in (b), the low-pass and high-pass filters are found. On the other hand, in (c), the notch filter and the offset compensator are exemplified.
Figure 1
Figure 1
Electrical diagram of the complete EMG acquisition system. (a) The instrumentation amplifier is shown with the operational amplifier in integrating configuration, while in (b), the low-pass and high-pass filters are found. On the other hand, in (c), the notch filter and the offset compensator are exemplified.
Figure 2
Figure 2
View (a) of the first floor of the EMG acquisition system where the two acquisition channels and the power supply system are located, and (b) the second floor of the EMG acquisition system where the two remaining channels and the acquisition system are located.
Figure 3
Figure 3
Window with the 4 graphs for each channel in the user interface.
Figure 4
Figure 4
Flow chart for the integration of genetic algorithms and SVM for feature selection algorithm.
Figure 5
Figure 5
Original, filtered, and surround signal from channel 1, person 1, and test 1 for the toe-off movement.
Figure 6
Figure 6
The original, filtered, surround signal from channel 2, person 1, and test 1 for arm extension movement.
Figure 7
Figure 7
From top to bottom, windows 40, 41, 42, 43, and 44 of the signal of channel 1 of the person 1 of the repetition 1 of the EMG signals of the right leg are represented.
Figure 8
Figure 8
Box plot of the absolute value characteristic corresponding to the seven movements of channel one of person two (a) considering the unsegmented signal, (b) considering window 20, (c) considering window 40, and (d) considering window 64.
Figure 9
Figure 9
Box plot of absolute value characteristic corresponding to the seven movements of channel one of person eight (a) considering the unsegmented signal, (b) considering window 20, (c) considering window 40, (d) considering window 64.
Figure 10
Figure 10
Plot of the classification percentage as a function of the number of iterations for (a) the Gaussian kernel and (b) the linear kernel.
Figure 11
Figure 11
Results of the average error percentage for the training and validation of the final SVM with (a) Gaussian kernel and with a (b) Linear kernel.
Figure 12
Figure 12
Results of the average error percentage for the training and validation of the final SVM with (a) Gaussian kernel and with a (b) Linear kernel.
Figure 13
Figure 13
Results of the average error percentage for the training and validation of the final SVM with (a) Gaussian kernel and with (b) linear kernel.
Figure 14
Figure 14
Graph of the classification percentage as a function of the number of iterations using the Gaussian kernel for (a) repetition 2 and (b) repetition 3.
Figure 14
Figure 14
Graph of the classification percentage as a function of the number of iterations using the Gaussian kernel for (a) repetition 2 and (b) repetition 3.
Figure 15
Figure 15
Plot of the classification percentage as a function of the number of iterations for (a) the linear kernel and (b) the Gaussian kernel.
Figure 16
Figure 16
The number of features given by PSO using the SVM error as a fitness function for the various configurations.
Figure 17
Figure 17
Plot of the classification percentage as a function of the number of iterations for (a) the linear kernel and (b) the Gaussian kernel.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Fang C., He B., Wang Y., Cao J., Gao S. EMG-centered multisensory based technologies for pattern recognition in rehabilitation: State of the art and challenges. Biosensors. 2020;10:85. doi: 10.3390/bios10080085. - DOI - PMC - PubMed
    1. Toledo-Pérez D.C., Rodríguez-Reséndiz J., Gómez-Loenzo R.A., Jauregui-Correa J.C. Support Vector Machine-based EMG signal classification techniques: A review. Appl. Sci. 2019;9:4402. doi: 10.3390/app9204402. - DOI
    1. Chowdhury R.H., Reaz M.B.I., Ali M.A.B.M., Bakar A.A.A., Chellappan K., Chang T.G. Surface electromyography signal processing and classification techniques. Sensors. 2013;13:12431–12466. doi: 10.3390/s130912431. - DOI - PMC - PubMed
    1. Kehri V., Awale R. EMG Signal Analysis for Diagnosis of Muscular Dystrophy Using Wavelet Transform, SVM and ANN. Biomed. Pharmacol. J. 2018;11:1583–1591. doi: 10.13005/bpj/1525. - DOI
    1. Dhindsa I.S., Agarwal R., Ryait H.S. Performance evaluation of various classifiers for predicting knee angle from electromyography signals. Expert Syst. 2019;36:e12381. doi: 10.1111/exsy.12381. - DOI

LinkOut - more resources