A hybrid algorithm for artifact rejection in EEG recordings based on iterative ICA and fuzzy clustering.

Brain responses to repeated sensory stimuli are typically contaminated by extraneous activity, including background rhythms, artifacts, and interference signals. To address this issue, we have recently proposed a new iterative independent component analysis (iICA) approach that can provide reliable evoked response (ER) estimates on a single trial basis. In this paper, we present a new two-step approach that focuses on removing well-defined artifacts, such as eye movements and muscle activity, before iICA processing. Extended analyses with both simulated data and actual recordings from normal subjects demonstrate that this procedure gives better results than iICA alone. Additionally, this methodology is suitable for fast analysis of multi-electrode recordings in parallel architectures, as individual channels can be processed simultaneously on different processors, and thus, it may have a significant impact on the analysis efficiency of large datasets of single-trial ERs.