Dynamic Brain Sources of Single-Trial Auditory Evoked Potentials Data Using Complex ICA Approach
Liangyu Zhao, Jianting Cao, Tetsuya Hoya, Andrzej S Cichocki · 2005
In this study, a robust pre-whitening technique and independent component analysis (ICA) approach are applied to unaveraged single-trial multichannel EEG data from auditory evoked potential (AEP) experiments. Single-trial event-related potential (ERP) data are usually averaged firstly to analysis in order to increase their signal/noise in electroencephalographic (EEG) experiment. However, averaging ignores the trial-by-trial variation of the amplitude. Our approach is based upon the two techniques: decorrelation with a highlevel additive noise reduction and decomposition of individual source components. The results on the unaveraged auditory evoked potential single-trial data analysis illustrate that not only the behavior and location but also the activity strength (amplitude) and dynamics of the individual evoked response can be visualized by the proposed method.