Detection Depth of Sleep in EEG Signal Using Hjorth Descriptor by Detecting NREM Condition

Abstract

Sleep is one of resting condition which is marked by a decrease in physical activity, awareness, and respond from external stimulation. Sleep is important in human’s life for regaining energy, maintain body resilience, and cognitive function of the body. Sleep quality can be seen by observing two conditions, the Rapid Eye Movement (REM) and Non-Rapid Eye Movement (NREM). These two conditions should be experienced by someone in order to gain a deep sleep state. Not experiencing one of them may affect to a negative disturbance in the body. This study has succeeded in finding a deep sleep state by detecting NREM condition in EEG signals which recorded during sleep. This study specifically analyzed delta and theta wave and used Hjorth Descriptor to extract the signal features. The features then fed to linear support vector machine (SVM) to detect normal and medicated deep sleep condition.