The automatic detection of slow eye movements (SEMs): A new method for sleep onset detection

Introduction: Despite the fact that excessive daytime sleepiness (EDS) has a significant impact on public health, we still lack objective measurements of EDS. A characteristic feature of the wake-sleep transition is the presence of slow eye movements (SEMs). We devised an algorithm for the automatic detection of SEMs on the electrooculogram. Objective: Validate the algorithm performance; compare the performance of automatic detection of SEMs versus the standard measurements of MSLT; perform the automatic SEMs detection on both the MSLT and the maintenance wakefulness test (MWT) recordings in a set of patients with Obstructive Sleep Apnea Syndrome (OSAS). Conclusion: Our algorithm is a reliable tool for automatic SEM detection. SEMs can be easily detected automatically and represent an effective marker of sleepiness in those conditions usually characterized by sleep onset with NREM sleep. Furthermore, automatic SEM detection was comparable to the standard polysomnographic assessment of sleep onset, thus providing a simplified technical requirement for the MSLT and the MWT. Further studies are warranted to evaluate SEM detection in other sleep disorders.