Pre-sleep/Sleep/Post-sleep stages

Abstract: Sleep disorders are treated as one of the major issues in human life. Therefore, automatic and effective system which can segregate the sleep stages and duly assist physicians and clinicians to diagnose the sleep related disorders is very much in demand.  Body movements as well as responsiveness towards any sort of external stimuli are consequently declined. Deep sleep or slow wave sleep plays a vital role in the sleep process, during which the body retrieves from its exhaustion. The proposed work focused on developing a less complex sleep stage classification algorithm which functions quicker by analysing single EEG channel.  By adopting the Empirical Mode Decomposition of Electro encephalography (EEG) signals, Huang -Hilbert Transform (HHT) the features namely i) Kurtosis, ii) IQR and iii) MAD were extracted which decipher in various sleep signals. We intend to deploy this feature as an important element in automatic classification of sleep stages namely pre-sleep, awake and post sleep stages. The adopted system attained an accuracy of 92% which is potential in classifying the sleep stages, which is major part in diagnosing the sleep related disorders.