Human Brain Mapping 2002
Sendai, Japan
June 10-16, 2002
Tzyy-Ping Jung, Erik Visser, Marissa Westerfield, Scott Makeig, Terrence J. Sejnowski
Institute of Neural Computation, University of California, San Diego, La Jolla, CA & Computational Neurobiology Lab., the Salk Institute for Biology Studies, La Jolla, CA {jung, visser, marissa, scott, and terry} @salk.edu
Interrelationship between P300 and alpha rhythms in single trials of a visual selective attention experiment
Introduction: The study of human event-related brain dynamics has been dominated by the study of averaged event-related potentials (ERPs) and fields (ERFs) recorded and averaged time locked to experimental events. Response averaging largely ignores the fact that the response waveforms may vary widely across trials and, more importantly, interactions between ERPs and ongoing EEG. This study analyzes unaveraged single-trial ERP data to examine interactions between ongoing EEG activity and the late positive complex (P300) of the visual target-response ERP.
Methods: Fifteen right-handed subjects, eight males and two females participated in a visual spatial selective attention task. During 72-second trial blocks, subjects were instructed to attend to one of five (1.6 cm) squares displayed horizontally at angles of 0X, 2.7X and 5.5X in the visual field 2X above from the point of fixation. Four squares were outlined in blue while one, marking the attended location, was outlined in green. Target stimuli were filled circles flashed at the attended location, and subjects were required to press a button as soon as possible following target stimulus presentation. Independent Component Analysis (ICA), a blind source separation technique, was performed on the concatenated 600+ single 1-sec target-centered EEG trials from each session to extract maximally independent components arising from different brain or extra-brain networks. For details, see Jung et al. (Human Brain Mapping, 14(3), 2001), Makeig et al. (Science, 2002).
Results: The derived independent components represented contributions from eye and muscle artifacts and from brain networks producing ongoing EEG and ERP activity. Examining single-trial activations of response-locked independent components made it possible to classify subtypes of ERP trials by reducing confounds from large artifacts and non-task related background EEG activity. For each subject and trial, activity of the components accounting for most of the averaged P300 (P3/P3b) were back-projected onto and summed at each scalp channel. For every subject, in 10-30% of the trials the resulting P300-dominated data did not resemble the averaged P300-component activation. In these 'inconsistent-P3' trials, although the subject responded correctly to the target stimulus, the 'P300' appeared absent. Latencies of early ERP peaks (P1 and N1) did not vary with subject reaction time (RT), and their amplitudes did not differ between consistent-P3 and inconsistent-P3 trials. However, in the singled inconsistent-P3 trials the post-stimulus phase distribution of lateral posterior alpha activity was less evenly distributed than in consistent-P3 trials. This enhanced and prolonged stimulus-induced phase resetting in the inconsistent-P3 trials, revealed by ICA, was very consistent across subjects. Scalp topographies of the affected components generally resembled projections of single equivalent current dipoles in lateral occipital cortex. Larger independent sources of posterior alpha activity, projecting to central posterior scalp, and right and left sensori-motor or mu rhythms, however, showed no such phase reset difference.
Conclusion: ICA can separate machine noise, physiological artifacts, event-related activities, and background EEG activities into different independent components, allowing us to examine functionally independent brain processing systems. The interrelationship between P300 amplitude and the event-related dynamics of lateral posterior alpha activity clearly demonstrates an interaction, uncovered by single-trial analysis, between ERPs and the ongoing EEG, contrasting with the common assumption that the averaged ERP evoked by brief visual stimuli reflects neural activity wholly separate from 'background' EEG.