Quantitative analysis of 'organization' by feature extraction of the EEG power spectrum

Quantitative analysis of 'organization' of the background EEG activity was made by applying a new method of approximating the optimal intensity-fitting function to the EEG records of healthy college students. The parametric intensity-fitting function used in this study to approximate the power spectrum of EEG was a gaussian type function. Among the quantitative parameters obtained by the new method, the ratio of the integrated power of alpha rhythm to that of all rhythms and the standard deviation of the alpha frequency proved to be the most important organization parameters by inspecting the Pearson product-moment correlation coefficient. Next the automatic scoring of 'organization' was successfully performed by using the linear regression of those two organization parameters.