Automatic detection of EEG arousals by use of normalized parameters for different subjects

Recent years, increase of patients with a sleep apnea syndrome (SAS) becomes a serious problem. The EEG arousals appeared in the sleep EEG of whole night recording gives specific information for clinical diagnosis on sleep disorders. In this paper, the automatic detection of EEG arousals in PSG record was proposed. In order to restrain the differences of characteristics of EEG between subjects, the normalization procedures were adopted for calculating parameters. Furthermore, the respiratory information of subject was also considered for determining the threshold values. The proposed method was applied to the data acquired from three subjects and the accuracy was investigated by comparing the visual inspection.