Dynamic scaling behavior of human brain electroencephalogram

After filtering out the α and β peaks in the power spectrum of the human brain electroencephalogram signals Y(t′), the probability distribution of the variation ΔY(t′)≡Y(t′+Δt)-Y(t′) exhibits a dynamic scaling behavior. The auto-correlation functions, persistence probabilities and detrended fluctuation functions of the time series |ΔY(t′)| all show a nontrivial power-law behavior. These results indicate that the dynamics of the human brain electroencephalogram is long-range correlated in time.