SCCN Home

Trends in Cognitive Science
May, 2004

Makeig S, Debener S, Onton J, Delorme A, "Mining Event-Related Brain Dynamics." Trends in Cognitive Science, 8(5):204-210, 2004.

A new method for analyzing high-density EEG and/or MEG data uses trial-by-trial visualization and time/frequency analysis to model the event-related dynamics of many cortical areas that contribute distinctive information to the recorded signals.

Click on the cover image at left to download the authors' preprint (.pdf, 616k).

Click here to view the online publication.

This article provides a new, more comprehensive view of event-related brain dynamics founded on an information-based approach to modeling electroencephalographic (EEG) brain dynamics. Most EEG research focuses either on peaks 'evoked' in average event-related potentials (ERPs) or on changes 'induced' in the EEG power spectrum by experimental events. Though these measures are nearly complementary, they do not fully model the event-related dynamics in the data, and cannot isolate the signals of the contributing cortical areas. We propose that many ERP and other EEG features are better viewed as time/frequency perturbations of ongoing field potential processes. The new approach combines independent component analysis (ICA), time/frequency analysis, and trial-by-trial visualization to measure EEG source dynamics without requiring an explicit head model.