Investigation of Neonatal EEG Time Series Using a Modified Nonlinear Dynamical Analysis

摘要

The Grassberger-Procaccia algorithm for computation of the correlation dimension is widely used nonlinear dynamical analysis techniques for EEG time series analysis. Even though the correlation dimension D_2 is the easiest dimension to compute, major drawback of the Grassberger-Procaccia algorithm is its extensive computational requirements. To overcome this, we introduce a modified computational algorithm referred to as the partial correlation integral. The partial correlation integral algorithm provides an approximation of the correlation dimension referred to as the dimensional exponent. Similar to the correlation dimension, the dimensional exponent can serve as a relative index of the complexity of a nonlinear dynamical system. In this study, the partial correlation integral algorithm is applied to analyze neonatal EEG sleep data. From the computational results, conclusions consistent with those made in previous studies using the correlation dimension are obtained.