Classification of raw myoelectric signals using finite impulse response neural networks

Author

Atsma, W.J. ; Hudgins, B. ; Lovely, F.

Author_Institution

Inst. of Biomed. Eng., New Brunswick Univ., Fredericton, NB, Canada

Volume

4

fYear

1996

fDate

31 Oct-3 Nov 1996

Firstpage

1474

Abstract

A method for classifying movement patterns of the upper arm, intended for multifunction control of arm prostheses, is presented. A finite impulse response neural network (FIRNN) is trained on 100 msec segments of myoelectric signals (MES) recorded during the very initial stage of elbow flexion (FL) and extension (EX). The network develops a clear internal representation of the input signals and is capable of classifying them

Keywords

artificial limbs; biomechanics; electromyography; medical signal processing; neural nets; 100 ms; arm prostheses control; elbow flexion; extension; finite impulse response neural network; input signals representation; movement patterns classification; multifunction control; myoelectric signal segments; raw myoelectric signals classification; upper arm; Biomedical engineering; Delay; Elbow; Electrodes; Multi-layer neural network; Multilayer perceptrons; Muscles; Neural networks; Neural prosthesis; Neurons;

fLanguage

English

Publisher

ieee

Conference_Titel

Engineering in Medicine and Biology Society, 1996. Bridging Disciplines for Biomedicine. Proceedings of the 18th Annual International Conference of the IEEE

Conference_Location

Amsterdam

Print_ISBN

0-7803-3811-1

Type

conf

DOI

10.1109/IEMBS.1996.647511

Filename

647511