Surrogate data modeling the relationship between high frequency amplitudes and Higuchi fractal dimension of EEG signals in anesthetized rats

Abstract

We used spectral analysis and Higuchi fractal dimension (FD) to correlate the EEG spectral characteristics of the sensorimotor cortex, hippocampus, and pons with their corresponding EEG signal complexities in anesthetized rats. We have explored the quantitative relationship between the mean FDs and EEG wide range high frequency (8-50 Hz) activity during ketamine/xylazine versus nembutal anesthesia at surgical plane. Using FD we detected distinct inter-structure complexity pattern and uncovered for the first time that the polygraphically and behaviorally defined anesthetized state at surgical plane as equal during experiment in two anesthetic regimens, is not the same with respect to the degree of neuronal activity (degree of generalized neuronal inhibition achieved) at different brain levels. Using the correlation of certain brain structure EEG spectral characteristics with their corresponding FDs, and the surrogate data modeling, we determined what particular frequency band contributes to EEG complexities in ketamine/xylazine versus nembutal anesthesia. In this study we have shown that the quantitative relationship between higher frequency EEG amplitude and EEG complexity is the best-modeled by surrogate data as a 3rd order polynomial. On the base of our EEG amplitude/EEG complexity relationship model, and the evidenced spectral differences in ketamine versus nembutal anesthesia we have proved that higher amplitudes of sigma, beta, and gamma frequency in ketamine anesthesia yields to higher FDs.