MULTICLASS CLASSIFICATION METHOD FOR THE ESTIMATION OF EEG SIGNAL QUALITY

摘要

The present invention relates to a method for assessing the quality of an electroencephalographic signal (EEC) based on a multiclass classification, wherein said method comprises the following steps: receiving (REC) at least one segment of electroencephalographic signal (S) acquired from at least one electrode; extracting (EXT) at least one feature value (F) from each channel of the electroencephalographic signal segment (S); classifying with a first classification (CLAS1) so as to assign each channel of the electroencephalographic signal segment (S) to one of at least three quality classes (TAG): {TAG1, TAG2,..., TAGN}; wherein said first classification (CLAS1) is performed by a k-nearest neighbors' algorithm.