On the use of high-order cumulant and bispectrum formuscular-activity detection

EUGENIO OROSCO; PABLO DIEZ; ERIC LACIAR; VICENTE MUT; CARLOS M SORIA; FERNANDO DISCIASCIO

BIOMEDICAL SIGNAL PROCESSING AND CONTROL

ELSEVIER SCI LTD

Lugar: Amsterdam; Año: 2015

1746-8094

The electromyographic (EMG) signals are extensively used on feature extraction methods for move-ment classification purposes. High-order statistics (HOS) is being employed increasingly in myoelectricresearch. HOS techniques could be represented in the frequency domain (high-order spectra, e.g., bis-pectrum, trispectrum) or in the time domain (higher-order cumulants). More calculus is required forcomputing the HOS in the frequency domain. On the one hand, classical bispectrum-based features wereapplied to EMG signals. We propose novel third-order cumulant-based features for EMG signals. Threedifferent classifiers are implemented for muscular-activity detection. Different analysis and evaluationswere applied to both HOS-based features in order to qualify and quantify similarities. Based on theseresults, it is possible to conclude that cumulant-based features and bispectrum-based features had com-parable behavior and allowed similar classification rates. Hence, extra calculus in order to convert time-to frequency-domain should be avoided.