New application of the Z estimator to identify the cognitive P300 in non-averaged human brain potentials.

Electronic averaging is currently used for displaying event-related potentials such as P300 from the electroencephalogram (EEG). However, there is a growing need for upgraded methods allowing cognitive components to be identified in single trial brain responses. The Z estimation method has been adapted for the topographic testing of non-averaged scalp recordings. Z values approximating +1 help validate a genuine P300 while failure to pass the Z test may suggest the spurious nature of a late positivity that mimicks a P300. Z testing can also identify interference from transient EEG alpha activity by showing alternations between +1 and -1 values, as expected from an oscillating alpha generator. In contrast with previous methods based on EEG recordings from a single scalp site, our topographic Z method takes into account 28 scalp sites for single trials testing.