Automatic Determination of Sleep-Wake States from EOG Signals using Fusion Technique

This paper deals with an alternative method in contrast to common classifiers that use very complex methods such as statistical pattern recognition, fuzzy system and neural networks for determining sleep-wake states from the Electrooculogram signals (EOG). The method involves a straight forward EOG based signal processing and feature extraction algorithm based on Lissajous-like plots. This novel method will be presented along with a case study to show its performance.