Ocular artifact removal from EEG: A comparison of subspace projection and adaptive filtering methods

One of the fundamental challenges in EEG signal processing is the selection of a proper method to correct ocular artifacts in the recorded electroencephalogram (EEG). Several methods have been proposed for this task. Among these methods, two main categories, namely subspace projection and adaptive filtering, have gained more popularity and are widely used in EEG processing applications. The main objective of this paper is to perform a comparative study of the performances of these methods using two measures, namely the mean square error (MSE) and the computational time of each algorithm. According to this study, ICA (independent component analysis) methods appear to be the most robust but not the fastest ones. Hence, they could be easily used for off-line applications. Moreover, PCA (principal component analysis) is very fast, but less accurate, so it could be used for real-time applications. Finally, adaptive filtering appears to have the worst performance in terms of accuracy, but it is very fast. Therefore, it could be also used for real-time applications, in which speed matters more than accuracy.