Wavelet-based fractal analysis of multi-channel epileptic ECoG

In this work, the spectral exponent y derived from the wavelet-based representation for 1/f processes is used to analyze multi-channel electrocorticogram (ECoG) data obtained from a subject with temporal lobe epilepsy. The computational results show that the spectral exponents of different channels of the ECoG data exhibit different characteristics of spectral exponent y. Also during an epileptic seizure the spectral exponent y of the ECoG signals is significantly higher than that associated with other states of the brain. An increase in the spectral exponent can however be observed in only channels that exhibit the ictal behavior, suggesting that regions of the brain that are involved in ictal activity can be localized.