Lempel-Ziv 복잡성 척도들의 근 피로도 추정 성능 평가

초록

The purpose of this study is to compare the muscle fatigue estimation performance of three complexity measures, LZ(Lempel-Ziv), PLZ(Permutation LZ) and DLZ(Dispersion LZ) with the conventional frequency measure (MDF:median frequency) based on robustness and sensitivity. Surface EMG(electrmyogram) signals(a total of 198 signals) were recorded in biceps brachii muscle with isometric 20%, 50% and 80%MVC(maximal voluntary contraction)s from eleven subjects and then, in this study for the first time, DLZ measure was applied to estimate muscle fatigue from these signals. Four measures were calculated from the 198 SEMG signals and the resulting fatigue estimation performance was evaluated in the point of robustness and sensitivity. Results show that DLZ measure with adopting four symbolic sequence(   ) is appropriate for muscle fatigue estimation and the conventional MDF measure is more robust (over mean 12%) and sensitive(over mean 88%) than three LZ complexity measures(LZ, PLZ, DLZ) for estimating muscle fatigue from surface EMG signal.

키워드