Fuzzy EMG classification for prosthesis control

TLDR: The fuzzy approach to classify single-site electromyograph (EMG) signals for multifunctional prosthesis control is superior to an artificial neural network method in at least three points: slightly higher recognition rate; insensitivity to overtraining; and consistent outputs demonstrating higher reliability.

Abstract: Proposes a fuzzy approach to classify single-site electromyograph (EMG) signals for multifunctional prosthesis control. While the classification problem is the focus of this paper, the ultimate goal is to improve myoelectric system control performance, and classification is an essential step in the control. Time segmented features are fed to a fuzzy system for training and classification. In order to obtain acceptable training speed and realistic fuzzy system structure, these features are clustered without supervision using the Basic Isodata algorithm at the beginning of the training phase, and the clustering results are used in initializing the fuzzy system parameters. Afterwards, fuzzy rules in the system are trained with the back-propagation algorithm. The fuzzy approach was compared with an artificial neural network (ANN) method on four subjects, and very similar classification results were obtained. It is superior to the latter in at least three points: slightly higher recognition rate; insensitivity to overtraining; and consistent outputs demonstrating higher reliability. Some potential advantages of the fuzzy approach over the ANN approach are also discussed.

Figures

TABLE I TYPICAL EMG CLASSIFICATION SYSTEMS

TABLE IV ERRORRATE CAUSED BY OVER-TRAINING WITH B DEFINED IN TEXT

Fig. 1. EMG classification strategy by ANN approach.

Fig. 2. Basic configuration of fuzzy logic system with fuzzifier and defuzzifier.

Fig. 3. Network representation of fuzzy logic system( fz is the gaussian fuzzfier, wherej = 1; 2; . . . ;M; i = 1; 2; . . . ; n. See text for details).

Fig. 4. Schematic diagram of fuzzy approach for EMG classification (MAV1 through WCare the 24 segmented features extracted; the initialization of fuzzy sets in fuzzy logic system is activated only once with clusters generated from Basic ISODATA algorithm; fuzzy logic system is trained and generates four outputs representing the possibility of each function; Output logic makes choice among these four outputs.)