EEG Signals processing two state discrimination using self-organizing maps

At present, there are many reasons why persons are affected in their ability to communicate with the society, so it is necessary to find an alternative communication channel for these people. The primary objective of this work is to process electroencephalographic (EEG) signals related to two specific mental task; which are also used to give Yes/No type short answers using signals produced by the brain. These signals come from two electrodes placed directly over the scalp. Obtained signals are related to specific commands or motion intention which can be used to generate an interaction channel for people-who have lost their standard capabilities of communication-with the society. Different processing methods for EEG signals were implemented, analyzed and classified in state of the art. In this paper, a Kohonen self-organizing map is proposed as the classifier. The obtained results give errors of 6% to 7%. The data used in this work was taken from the database of Universidad Peruana Caytano Heredia.