ESTIMATION OF EVENT-RELATED SYNCHRONIZATION CHANGES BY A NEW TVAR METHOD

The modeling of nonstationary electroencephalogram (EEG) with time-var ying autoregressive (TVAR) models is discussed, The classical least sq uares TVAR approach is modified so that prior assumptions about the si gnal can be taken into account in an optimal way, The method is then a pplied to the estimation of event-related synchronization changes in t he EEG, The results show that the new approach enables effective estim ation of the parameter evolution of the time-varying EEG with better t ime resolution compared to previous methods, The new method also allow s single-trial analysis of the event-related synchronization.