Abstract
Classic coherence analysis can detect coordination of EEG rhythms between brain areas. However, it could not reflect the dynamic properties changing with time. Time-varying coherence analysis is a method developed on base of classic coherence analysis and signal?s joint time-frequency representations in recent years. It was used to extract transient characteristics of the interactions among brain areas and describe the temporal, spatial and frequency relationships of brain activities. This paper discussed the probability of applying time-varying coherence to event related EEG to study the coordination mechanism of brain. Aimed at the applications of time-varying coherence analysis methods based on non-parameter estimation and multi-variant parameter model to EEG, we constructed the simulated EEG data, looked for the suitable approach for cognitive EEG and tried to apply it to the EEG data from classic Stroop effect. The results showed that appropriate time-varying analysis methods were effective tools for event related EEG to study the coordination between brain areas.