Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Three machine learning techniques for automatic determination of rules to control locomotion.

S Jonić1, T Janković, V Gajić

  • 1Faculty of Electrical Engineering, University of Belgrade, Yugoslavia.

IEEE Transactions on Bio-Medical Engineering
|March 31, 1999
PubMed
Summary

Three machine learning techniques predict muscle activation for functional electrical stimulation (FES) assisted walking. Adaptive-network-based fuzzy inference system (ANFIS) generated the most interpretable rules for FES control.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Novel surgical procedure in design of advancement endorectal flap in the surgical treatment of complex anal fistulas.

Rozhledy v chirurgii : mesicnik Ceskoslovenske chirurgicke spolecnosti·2024
Same author

Scipion Flexibility Hub: an integrative framework for advanced analysis of conformational heterogeneity in cryoEM.

Acta crystallographica. Section D, Structural biology·2023
Same author

ScipionTomo: Towards cryo-electron tomography software integration, reproducibility, and validation.

Journal of structural biology·2022
Same author

Association between cardiac high-energy phosphate metabolism and whole body metabolism in healthy female adults.

Physiological research·2021
Same author

Gallbladder carcinoma - when to use laparoscopic approach?

Bratislavske lekarske listy·2021
Same author

The implications of hypersomnia in the context of major depression: Results from a large, international, observational study.

European neuropsychopharmacology : the journal of the European College of Neuropsychopharmacology·2019

Area of Science:

  • Biomedical Engineering
  • Robotics
  • Machine Learning

Background:

  • Accurate prediction of gait events and muscle activation is crucial for real-time locomotion control.
  • Functional electrical stimulation (FES) assisted walking requires precise control of muscle activations.
  • Supervised machine learning (ML) offers potential for predicting muscle activation patterns.

Purpose of the Study:

  • To present and compare three supervised ML techniques for predicting muscle activation patterns during FES-assisted walking.
  • To evaluate ML models based on their ability to generate rules for FES controllers, generalization, computational complexity, and learning rate.
  • To predict knee flexor muscle activation and knee joint angle for FES-assisted walking.

Main Methods:

  • Three ML techniques were investigated: 1) Multilayer perceptron (MLP) with Levenberg-Marquardt, 2) Adaptive-network-based fuzzy inference system (ANFIS), and 3) Inductive learning (IL) combined with a radial basis function (RBF) network.

Related Experiment Videos

  • Models were trained and tested using simulated data of FES-assisted walking, focusing on knee joint angle and ground reaction forces.
  • Performance was assessed by the interpretability and number of generated rules, generalization, computational complexity, and learning speed.

    • Main Results:

    • ANFIS produced the minimal, most explicit, and comprehensible rules, ideal for FES controller generation.
    • The combination of IL and RBF network demonstrated the best generalization capabilities.
    • Predictions were made for knee flexor muscle activation and knee joint angle over seven consecutive strides.

    Conclusions:

    • ML techniques show promise in predicting muscle activation for FES-assisted locomotion.
    • ANFIS is advantageous for its rule interpretability in FES control applications.
    • The IL-RBF combination offers superior generalization for gait prediction in FES-assisted walking.