Linear automatic detection of eye movements during the transition between wake and sleep

Slow eye movement (SEM) are a sensitive indicator of lowered consciousness or drowsiness in man. A new computerized method for detection of SEM was introduced. A linear regression analysis was applied in each moving window for approximation to the tangent line on the electro-oculogram curve. The results revealed that SEM were more frequent and their duration was shorter at stage wake than at sleep stages 1 and 2. The method was practically suitable for objective detection and measurement of SEM.