Hilbert-Huang transform based classification of sleep and wake EEG signals using fuzzy c-means algorithm

Sleep is very essential for physical and mental health of human. With the enlargement of modernity, a huge amount of the world population is suffering from many sleep disorders. These disorders are able to influence patient´s physical, mental, emotional and social functioning. In this paper, the Hilbert-Huang transform (HHT) based features are extracted for classification of sleep and wake EEG signals. HHT decomposes non-stationary and nonlinear signal into a sum of analytic intrinsic mode functions (IMFs). The bandwidth features namely amplitude modulation bandwidth (B_AM) and frequency modulation bandwidth (B_FM) are extracted from analytic representation of IMFs. These parameters are given in fuzzy clustering unsupervised learning technique using fuzzy c-means (FCM) algorithm. The experimental outcomes shows that the presented method is effective for classification of sleep and wake EEG signals using single IMF.