Evaluation of an automatic ocular filtering method for awake spontaneous EEG signals based on independent component analysis

Electroencephalographic artifacts associated with eye movements are a potential source of error in the EEG analysis when its interpretation is performed for evaluating the influence of drugs and the diagnosis of neurological disorders. In this study, a new automatic method for artifact filtering based on independent component analysis (ICA) is proposed. Automatic artifact identification is based on frequency domain and scalp topography aspects of the independent components. A comparative study between ICA and the 'gold standard' method based on linear regression analysis is performed. The latter does not take into account the mutual contamination between EEG and electrooculographic activity, reducing not only the ocular movements but also interesting cerebral activity, mainly in anteriorly placed electrodes. This limitation is overcome by ICA and the efficiency of this approach is shown for a double-blind, placebo-controlled crossover drug trial in healthy volunteers.