Wavelet-based fractal analysis of sleep EEG

Temporal patterns of the EEG provide insight into the underlying states of the brain. The spectral exponent γ derived from the wavelet-based representation for 1/f processes is used to investigate the self-similarity of EEG during sleep. An increase in the spectral exponent γ leads to sample signals with smoother temporal patterns. From the computational results, it is observed that the spectral exponent γ has distinguishable characteristics corresponding to different sleep stages. Further, the spectral exponent γ of the sleep EEG associated with a deeper sleep stage is significantly higher than that associated with a lighter sleep stage. This thus suggests that the wavelet-based fractal analysis is potentially a useful computational tool for quantifying sleep stages.