Modeling Nonstationary Time Series and Inferring Instantaneous Dependency, Feedback and Causality: An Application to Human Epileptic Seizure Event Data

Abstract The parametric modeling of nonstationary covariance time series and the determination of the instantaneously changing structure of the interdependencies between those time series, inferred from the fitted model are treated. The nonstationary time series are modeled by a multivariate time varying autoregressive (AR) model. The time evolution of the AR parameters are expressed as linear combinations of discrete Legendre orthogonal polynomial functions of time. The model is fitted by a Householder t|ransformat ion least squares-Akaike AIC order determination, regression subset selection method. The computation of the instantaneous dependence, feedback and causality structure of the time series, from the fitted model, is discussed. An example of the modeling and determination of instantaneous causality in a human implanted electrode seizure event EEG is shown.

[1]  N. Kawabata.  A nonstationary analysis of the electroencephalogram. , 1973, IEEE transactions on bio-medical engineering.

[2]  J. Geweke,et al.  Measurement of Linear Dependence and Feedback between Multiple Time Series , 1982 .

[3]  Peter E. Caines,et al.  Feedback between stationary stochastic processes , 1975 .

[4]  Walter Do,et al.  The method of complex demodulation. , 1968 .

[5]  H. Akaike.  Stochastic theory of minimal realization , 1974 .

[6]  Brian D. O. Anderson,et al.  Identifiability of linear stochastic systems operating under linear feedback , 1982, Autom..

[7]  G. Kitagawa,et al.  A smoothness priors time-varying AR coefficient modeling of nonstationary covariance time series , 1985, IEEE Transactions on Automatic Control.

[8]  W. Gersch.  Causality or driving in electrophysiological signal analysis , 1972 .

[9]  S. Bressler,et al.  Shadows of thought: shifting lateralization of human brain electrical patterns during brief visuomotor task. , 1983, Science.

[10]  Arthur C. Sanderson,et al.  Detecting change in a time-series (Corresp.) , 1980, IEEE Trans. Inf. Theory.

[11]  W. Gersch,et al.  Epileptic Focus Location: Spectral Analysis Method , 1970, Science.

[12]  A. Benveniste,et al.  Detection of abrupt changes in signals and dynamical systems : Some statistical aspects , 1984 .

[13]  H. Akaike.  A new look at the statistical model identification , 1974 .

[14]  Lennart Ljung,et al.  Identification of processes in closed loop - identifiability and accuracy aspects , 1977, Autom..

[15]  T. Bohlin.  Four Cases of Identification of Changing Systems , 1976 .

[16]  P. Crandall,et al.  Surface and Deep EEG Correlates of Surgical Outcome in Temporal Lobe Epilepsy , 1981, Epilepsia.

[17]  G. Bodenstein,et al.  Feature extraction from the electroencephalogram by adaptive segmentation , 1977, Proceedings of the IEEE.

[18]  J. Geweke,et al.  Measures of Conditional Linear Dependence and Feedback between Time Series , 1984 .

[19]  K Shinosaki,et al.  The dominant direction of interhemispheric EEG changes in the linguistic process. , 1981, Electroencephalography and clinical neurophysiology.

[20]  C. E. SHANNON,et al.  A mathematical theory of communication , 1948, MOCO.

[21]  M Hildebrand,et al.  Symmetrical gaits of horses. , 1965, Science.

[22]  C. W. J. Granger,et al.  Economic Processes Involving Feedback , 1963, Inf. Control..

[23]  A. Sarris,et al.  A Bayesian Approach To Estimation Of Time-Varying Regression Coefficients , 1973 .

[24]  F. D. Silva,et al.  Propagation of seizure activity in kindled dogs. , 1983 .