Proposed link rates in the human brain.

ABSTRACT

There is increasing experimental evidence that neuronal synchronization is necessary for the large-scale integration of distributed neuronal activity to realize various time-dependent coherent neuronal assemblies in the brain. Phase synchronization seems a promising candidate to quantify the time-dependent, frequency specific, synchrony between simultaneously recorded electroencephalogram (EEG) signals that may partially reflect this former process. We introduce a link rate (LR) as a measure of the spatial-temporal incidence of phase synchronization and phase de-synchronization. The concept is exemplified in its application to the analysis of spontaneous phase synchronization. To this end, three scalp EEG recordings are used a normal control, a patient suffering from epileptic seizures and a patient with diffuse brain damage due to anoxia, showing a burst-suppression EEG. In addition, the method is applied to surrogate data (white noise). We find in the normal control that LR(control)=13.90+/-0.04 (mean+/-S.E.M.), which is different from the surrogate data, where we find that LR(surr)=15.36+/-0.05. In the two pathological conditions, the LR is significantly and strongly reduced to LR(burst)=4.52+/-0.05 and LR(seizure)=5.40+/-0.08. The derived LR seems a sensitive measure to relevant changes in synchronization, as these occur in the dynamic process of generating different spatial-temporal networks, both in physiological and pathological conditions.