Astrid von Stein
The Neurosciences Institute, San Diego
In a series of EEG experiments we determined both, local and inter-areal synchronization in human subjects using power and crosspower analysis of EEG. We were able to demonstrate that during processing of simple visual features such as gratings the local degree of gamma power reflects the degree of synchronization between columns within a cortical area (von Stein et al.1995). We were further interested in the pattern of synchronization during higher cognitive functions. Presenting semantic stimuli such as daily life objects, we found an increased synchronization between temporal and parietal cortex, induced independent of the modality of presentation. The frequency range of this synchronized activity was 12-18 Hz, i.e. lower than the gamma range. Another significant increase of synchronized activity was induced during the rehearsal intervall of a visual working memory task, where very long range interactions appeared between prefrontal cortex and temporal cortex; the frequency range was even lower, 4-7 Hz (theta), in this case. We have observed similar long range cortico-cortical synchonization of slow frequency during several types of intrinsic mental activity such as mental imagery or mentally painting (e.g.Petsche et al. 1996), thus tentatively interpreting it as reflecting purely top-down driven synchronization of neuronal assemblies. Concluding, a relation between the spatial range of interaction and the frequency range of synchronization seems to be apparent in our results: local visual intergration showed synchronized activity on a fast temporal scale (i.e. gamma), multimodal integration between neighbouring temporal and parietal areas showed synchronization in an intermediate frequency (beta1), and mental activity needing very long-range interaction had synchronized activity within slow freuqncy ranges (i.e. alpha, or theta).
These findings seem evidence that both, the spatial extension of the involved network and the degree of complexity and time needed to integrate different levels might influence in which frequency range cortical dynamics evolves. Synchronization whithin different frequency ranges additionally opens the possibility that different cortical processes might coexist simultanously.