DETECTING FREQUENCY-SPECIFIC PHASE SYNCHRONY IN BRAIN ELECTRICAL OSCILLATIONS

Oscillations in brain electrical activity that occur synchronously across brain regions within specific frequency bands are of great interest in neuroscience. Detecting such phase synchrony in parallel recorded oscillatory signals like EEG or MEG requires some kind of narrow-band filtering as a preliminary step for the extraction of frequency-specific phase information. We address the ensuing technical problems of phase enslaving and spurious synchrony and propose a method designed to reduce these problems. The method is detailed for application to repeated trials aligned to stimulus onset. It is going to be used in an analysis of EEG correlates of endogeneous perceptual reversals induced by ambiguous stimuli. Synchronized neural activity is widely ascribed a major role in the functional organization of the brain. In object recognition, for example, various object features processed separately in distributed visual subsystems have to be integrated for a coherent percept to build up. Ac-cording to the binding-by-synchrony hypothesis this is accomplished by a common temporal signature of all neural network nodes participating in the representation of the object, namely by synchronous oscillations at specific frequencies. Such in-phase oscillations would selec