Abstract :
A unique statistical technique is described which appears to measure important interaction processes within the central nervous system (CNS). This statistic, termed ``weighted-average coherence,´´ or C¿, is derived primarily from parameters generated in cross-spectral frequency analysis. C¿ has the useful property, however, of summarizing a larger amount of cross-spectral information into a more manageable form. By applying C¿ analysis to the ongoing electroencephalogram (EEG), it has been possible to define unique patterns of interaction, or ``coupling,´´ between different brain areas. The total pattern of EEG coupling is taken to define a given state of functional brain organization. Moreover, since C¿ is sensitive to changes in the EEG, it also reflects something of the dynamic properties of such brain organization.
