| A study was undertaken to better characterize normal human spontaneous EEG and steady-state visually-evoked potentials (SSVEPs), and to relate these observations to theoretical models of local/global interactions within the neocortex. 131-channel scalp data were collected from normal subjects in relaxed and cognitive (mental addition) states. High priority was given throughout to data quality, artifact prevention, rigorous signal analysis, and extensive validation of methods.;Calculation of coherence at frequencies between 2 and 48 Hz showed mixed increases/decreases in alpha coherence, and increases in theta coherence with cognitive activity. Coherence patterns were systematically translated into valued and unvalued graphs and analyzed using tools from information theory. The resulting networks displayed high topological integration in cases of high total coherence. Comparison of graph distance with physical distance in these networks yielded evidence for binding of semi-global networks oscillating in the theta and upper alpha bands. Further study of phase information at peak alpha-band frequencies showed significant consistency of phase maps between epochs, consistent with existence of a globally resonant component in the alpha rhythm.;In order to further investigate these phenomena, methodology was developed and validated to accurately extract spatial spectra from high-density EEG. Frequency-wavenumber relations were identified in EEG and SSVEP, consistent with wavelike propagation of neural activity along the corticocortical fibers. Furthermore, changes observed during cognitive activity suggested a concomitant shift from global to local dynamics in some subjects.;Taken as a whole, these results indicate a role for global dynamics in the human alpha rhythm, and specifically suggest a mixed global/local origin for this phenomenon. The results also yield: strong evidence for wavelike propagation of the SSVEP throughout the neocortex; further evidence for a global-to-local transition during periods of increased cognitive activity; and a more complete baseline understanding of coherence in the normal human EEG. This, along with the methodological advances described, should provide a firm foundation for more effective use of scalp EEG in both clinical diagnosis and research applications. |
