Analysis of superimposed signals using time-scale phase representation

In this paper we address the decomposition problem of electromyographic signals (EMG) and we present a new idea for the decomposition approach introducing the time-scale phase representation. Considering the nature of EMG signals, a special set of nonorthogonal conjugated basis functions is used to perform the time-scale analysis. Mapping the modified signal phase into the time-scale phase plane is applied. Its minimization leads towards a very good time localization of individual superimposed signal components.