Detection of relevant information in intrinsic mode functions

The empirical mode decomposition (EMD) decomposes a local and adaptive time series into a finite set of intrinsic mode functions (IMF), AM-FM signals that allow to represent a non-linear and non-stationary model with the advantage of not losing the underlying meaning. This study examines time series of sEMG measurements for a case study of healthy individuals with carpal tunnel syndrome. Due to the amount of multiple levels of detail, all around a central frequency and evoked by the number of IMFs obtained through EMD, the informational contribution of each at the intermodal and interindividual level is studied through Shannon entropy to establish a general framework of spectral study given Hilbert Huang's (HHT) transformation to remarkable degrees of information. The results show that the latest IMFs have more disordered states even when they engage in apparently regular behavior, agglomerate more time-frequency information, and in the same way, concentrate more differentiable characteristics for a process of individualization of patterns.