In this study, a linear decomposition technique, independent component anal
ysis (ICA), is applied to single-trial multichannel EEG data from event-rel
ated potential (ERP) experiments. Spatial filters derived by ICA blindly se
parate the input data into a sum of temporally independent and spatially fi
xed components arising from distinct or overlapping brain or extra-brain so
urces. Both the data and their decomposition are displayed using a new visu
alization tool, the "ERP image," that can clearly characterize single-trial
variations in the amplitudes and latencies of evoked responses, particular
ly when sorted by a relevant behavioral or physiological variable. These to
ols were used to analyze data from a visual selective attention experiment
on 28 control subjects plus 22 neurological patients whose EEG records were
heavily contaminated with blink and other eye-movement artifacts. Results
show that ICA can separate artifactual, stimulus-locked, response-locked, a
nd non-event-related background EEG activities into separate components, a
taxonomy not obtained from conventional signal averaging approaches. This m
ethod allows: (1) removal of pervasive artifacts of all types from single-t
rial EEG records, (2) identification and segregation of stimulus- and respo
nse-locked EEG components, (3) examination of differences in single-trial r
esponses, and (4) separation of temporally distinct but spatially overlappi
ng EEG oscillatory activities with distinct relationships to task events. T
he proposed methods also allow the interaction between ERPs and the ongoing
EEG to be investigated directly. We studied the between-subject component
stability of ICA decomposition of single-trial EEG epochs by clustering com
ponents with similar scalp maps and activation power spectra. Components ac
counting for blinks, eye movements, temporal muscle activity, event-related
potentials, and event-modulated alpha activities were largely replicated a
cross subjects. Applying ICA and ERP image visualization to the analysis of
sets of single trials from event-related EEG (or MEG) experiments can incr
ease the information available from ERP (or ERF) data. (C) 2001 Wiley-Liss,
Inc.