Conditional linear random process and random coefficient autoregressive model for EEG analysis

Abstract

In the paper a new mathematical model of electroencephalographic (EEG) signal has been justified in the form of conditional linear random process (CLRP). The model has been derived from the biophysical nature of brain electrical activity. Random coefficient autoregressive (RCA) model (as a member of the class of discrete-time CLRP) has been applied for EEG signal processing. Using the special statistical test, RCA model has been shown to be more suitable for given EEG data than the autoregressive one with constant coefficients, especially for heteroscedastic EEG from epileptic patients.