INVESTIGATION OF EEG SIGNAL LENGTH INFLUENCE ON ACCURACY OF ANESTHESIA LEVELS CLASSIFICATION
This paper considers one of the challenging tasks during surgical procedure, i.e. depth of anasthesia estimate. The purpose of this paper is to investigate the effect of the analyzed EEG signal fragment duration on the accuracy of anesthesia level estimate using the linear discriminant analysis algorithm and determining the EEG signal length, which yields acceptable accuracy of anesthesia level separation using these parameters.A new method for classifying EEG anesthesia levels is proposed. The possibility of classifying levels of anesthesia is demonstrated by means of sharing the EEG parameters under consideration (SE, BSR, SEF95, RBR). The method can be used in anesthesia monitors that are used to monitor the depth of anesthesia in order to select the appropriate dose of anesthetic drugs during operations, thus avoiding both cases of intraoperative arousal and excessively deep anesthesia.