An information-theoretic approach suitable for studying synchronization phenomena in experimental time series has been applied in analysis of EEG recordings of epileptic patients. Transient phenomena leading to seizures have been characterized by increased synchronization (local and between areas) and asymmetry in information flow (the area of the epileptogenic focus drives and synchronizes adjacent areas).
Although the results should be regarded as preliminary, they suggest that the method has a promising potential for localization of epileptic foci and anticipation of approaching seizures.