Society for Neuroscience
Oct. 23-27, 2004 - 2004 - San Diego, CA
Functional significance of the novelty P3 event-related potential revealed by independent component analysis.
(1,2)S. Debener, S. Makeig(3), A. Delorme(3), A.K. Engel(1,2)
(1) Institute of Neurophysiology and Pathophysiology, University Hospital Hamburg-Eppendorf, Hamburg, Germany
(2) Institute of Medicine, Research Center Juelich, Juelich, Germany
(3) Swartz Center for Computational Neuroscience, Institute for Neural Computation, La Jolla, CA, USA
sdebener@uke.uni-hamburg.de
To better understand whether voluntary attention affects how the brain processes novel events, variants of the auditory novelty oddball paradigm were presented to two different groups of human volunteers. One group of subjects (n = 16) silently counted rarely presented 'infrequent' tones (p = 0.10), interspersed with 'novel' task-irrelevant unique environmental sounds (p = 0.10) and frequently presented 'standard' tones (p = 0.80). A second group of subjects (n = 17) silently counted the 'novel' environmental sounds, the 'infrequent' tones now serving as the task-irrelevant deviant events. Analysis of ERPs recorded from 63 scalp channels suggested a spatio-temporal overlap of fronto-central novelty-P3 and centro-parietal P3 (P3b) ERP features in both groups. Application of independent component analysis (ICA) to concatenated singletrials revealed two independent component clusters that accounted for portions of the novelty-P3 (FP3 cluster) and P3b (PP3 cluster) response features, respectively. The P3b-related ICA cluster PP3 contributed to the novelty-P3 amplitude response to novel environmental sounds. In contrast to the scalp ERPs, the amplitude of the novelty P3 related cluster (FP3) was not affected by voluntary attention, i.e., by the target/non-target distinction. This result demonstrates the usefulness of ICA for disentangling spatio-temporally overlapping ERP processes and provides strong evidence that task-irrelevance is not a necessary feature of novelty processing.