Electroencephalogram Signal Acquisition System with Machine Learning for Robotic Prosthesis Control: In Vivo Dataset

Authors

  • Bruno Santos Department of Electric, Federal Institute of São Paulo, Department of Information Technology, University São Judas Tadeu and Department of Electric, University of São Paulo, Brazil https://orcid.org/0000-0001-5176-4711
  • Alysson Avila Department of Electric, Federal Institute of São Paulo, Brazil
  • Marcelo Barboza Department of Electric, Federal Institute of São Paulo, Brazil
  • Evandro Drigo Department of Electric, Federal Institute of São Paulo, Brazil
  • Tarcisio Leão Department of Electric, Federal Institute of São Paulo, Brazil
  • Eduardo Bock Department of Electric, Federal Institute of São Paulo, Brazil

DOI:

https://doi.org/10.47852/bonviewAIA52024252

Keywords:

electroencephalogram, artificial intelligence, machine learning, ensemble AI, brain-machine interface, robotic upper limb prostheses

Abstract

This study addresses the increasing global need for upper limb prostheses (ULPs), particularly those controlled by electroencephalogram (EEG) signals, due to the rising number of amputations. Focusing on an EEG signal acquisition system integrated with a machine learning (ML)-driven pattern recognition framework, the research investigates the control of a robotic ULP. The study is divided into two phases: EEG data acquisition and pattern recognition using an ensemble of K-nearest neighbors (KNN), support vector machines (SVM), and artificial neural networks (ANN) models within a brain-machine interface. Each ML model demonstrated distinct strengths—KNN in rapid pattern recognition, SVM in reliable state differentiation, and ANN in handling complex, non-linear data relationships. The ensemble ML (eML) leveraged these strengths, achieving approximately 90% accuracy in final training rounds and showing superior performance compared to individual models. The eML was successfully integrated into the robotic ULP control system, demonstrating high potential for real-world applications by efficiently processing brain activity signals and making precise control decisions.

 

Received: 4 September 2024 | Revised: 25 October 2024 | Accepted: 7 November 2024

 

Conflicts of Interest

The authors declare that they have no conflicts of interest in this work.

 

Data Availability Statement

The data that support the findings of this study are openly available in GitHub at https://github.com/S-Brno/ARTIFICIAL_INTELLIGENCE__EEG_UPPERLIMB_PROSTHESIS.

 

Author Contribution Statement

Bruno J. Santos: Conceptualization, Methodology, Software, Validation, Formal analysis, Investigation, Writing – original draft, Visualization. Alysson Avila: Conceptualization, Software, Validation, Formal analysis, Investigation, Data curation. Marcelo Barboza: Writing – review & editing. Evandro Drigo: Resources, Writing – review & editing. Tarcisio Leão: Supervision. Eduardo Bock: Resources, Supervision, Project administration.


Metrics

Metrics Loading ...

Downloads

Published

2025-02-15

Issue

Section

Research Article

How to Cite

Santos, B., Avila, A., Barboza, M., Drigo, E., Leão, T., & Bock, E. (2025). Electroencephalogram Signal Acquisition System with Machine Learning for Robotic Prosthesis Control: In Vivo Dataset. Artificial Intelligence and Applications. https://doi.org/10.47852/bonviewAIA52024252