Exploratory EEG data analysis for neurophysiological experiments

A method for the exploratory analysis of (EEG) data for neurophysiological experiments is presented. It is based on a time-frequency decomposition of the EEG time series, which is measured at a number of electrodes in the skull surface, and includes: the computation of a statistic that measures the deviations of the log-power with respect to the pre-stimulus average; the computation of a significance index for these deviations, and the segmentation of the time-frequency plane into regions with uniform activation patterns. As particular example, an experiment to study EEG changes during figure and word categorization is analyzed in detail.