No abstract

No abstract

Signals obtained during tonic-clonic epileptic seizures are usually neglected for analysis by the physicians due to the presence of noise caused by muscle contractions. Although noise obscures completely the recording, some information about the underlying brain activity can be obtained with wavelet transform by filtering those frequencies associated with muscle activity. One great advantage of this method over traditional filtering is that the filtered frequencies do not modify the pattern of the remanent ones. An accurate analysis of the different seizure stages was achieved using the wavelet packet method, and through the information cost function the brain dynamical behavior can be accessed. ͓S1063-651X͑97͒09712-2͔

No abstract

No abstract

No abstract

No abstract

By recourse to appropriate information theory quantifiers (normalized Shannon entropy and Martín-Plastino-Rosso intensive statistical complexity measure), we revisit the characterization of Gaussian self-similar stochastic processes from a Bandt-Pompe viewpoint. We show that the ensuing approach exhibits considerable advantages with respect to other treatments. In particular, clear quantifiers gaps are found in the transition between the continuous processes and their associated noises.