A new regression-based method for the eye blinks artifacts correction in the EEG signal, without using any EOG channel.

Eye blinks artifacts correction in the EEG signal is a best practice in many applications. Nowadays, different approaches can be used to overcome such an issue: the most used methods are based on regression techniques and Independent Component Analysis. It is not clear which is the best performing method, thus the choice of which method to adopt depends on the specific application, on the basis of the method limitations. In fact, on one hand the regression-based methods require at least one EOG channel, and are affected by the mutual contamination between EEG and EOG signals. On the other hand, the ICA-based methods need a higher number of electrodes and a greater computational effort than the regression-based ones. In this study, a new regression-based method has been proposed and compared with three of the most used algorithms (Gratton, extended InfoMax, SOBI) for eye blinks correction. The results showed that the proposed algorithm was able (i) to achieve similar efficiency of the other methods in correcting the blinks, but without requiring neither EOG channels, nor a great electrodes number, nor a high computational effort, and (ii) to preserve EEG information in blink-free signal segments.