Single-trial multi-channel N170 estimation using linear discriminant analysis (LDA)

Abstract

N170 is one of the event-related potentials (ERPs) that have been extensively used to study the neurological response of a subject when presented with a visual face stimulus. Although many N170 experiments have been performed with multi-channel EEG recordings, most of the N170 analyses are still based on trial-averaged signals from one or a few selected electrodes. Not only does this method under utilise the information from all available electrodes, it inhibits a trial-to-trial analysis of N170. We address this issue by proposing a single-trial multi-channel N170 estimation method that estimates the latency, amplitude and scalp projection of N170 using a linear classifier-based approach. © 2012 Springer-Verlag.