EEG COHERENCY .1. STATISTICS, REFERENCE ELECTRODE, VOLUME CONDUCTION,LAPLACIANS, CORTICAL IMAGING, AND INTERPRETATION AT MULTIPLE SCALES

Several methodological issues which impact experimental design and phy siological interpretations in EEG coherence studies are considered, in cluding reference electrode and volume conduction contributions to err oneous coherence estimates. A new measure, 'reduced coherency', is int roduced as the difference between measured coherency and the coherency expected from uncorrelated neocortical sources, based on simulations and analytic-statistical studies with a volume conductor model. The co ncept of reduced coherency is shown to be in semi-quantitative agreeme nt with experimental EEG data. The impact of volume conduction on stat istical confidence intervals for coherence estimates is discussed. Con ventional reference, average reference, bipolar, Laplacian, and cortic al image coherencies are shown to be partly independent measures of ne ocortical dynamic function at different spatial scales, due to each me thod's unique spatial filtering of intracranial source activity. (C) 1 997 Elsevier Science Ireland Ltd.