Abstracts

Seizures may be the result of neuronal and network synchronization under certain conditions. Accurate and robust measures of synchronization in multi-channel nonstationary data are needed for in-depth probing of the role synchronization in ictiogenesis. We describe a new synchronization measure and provide a preliminary assessment of its performance., Average times of phase synchronization between multiple pairs of channels are calculated. These results are compiled to produce a synchronization-time matrix. A random-matrix based criterion is applied to determine the diagonal matrix elements and qualitative changes in the underlying system[apos]s dynamics are assessed by monitoring characteristic matrix quantities such as eigenvalues and the determinant. The method[apos]s performance was assessed using a control model of coupled oscillators and human EEG and ECoG data., The method accurately tracks changes in degree of coupling between oscillators. Seizures recorded from scalp and intracranially are associated with changes in degree of synchronization that are spatially inhomogeneous. Global measures such as determinant vary inter-individually and also intra-individually as a function of montage. Generalized seizures appear to be associated with global increase in phase synchronization., Seizures may be associated with increase or decreased synchronization and synchronization changes in opposing directions may occur simultaneously at different sites. The bases for this lack of homogeneity are under investigation. The apparent montage-dependence of these changes suggests that the process has a certain degree of spatial isotropy., (Supported by NIH/NINDS grant #5R01NS046602-02.)