FIGURE 1.

Schematics depicting validation of the statistical model. Only the eight most discriminative groups of electrodes defined from the training set were selected. Then, classifier training (CT) based on the reduced training set of n = 60 patients and classification (C) of the reduced validation set of m = 8 patients was performed in two approaches: (A) a single classifier trained on all n subjects from the training set; (B) voting of n classifiers trained on n–1 subsets of the training set.