Sleep apnea related micro-arousal detection with EEG analysis

Arousals are regarded as one of the most influential causes of daytime sleepiness and are frequent occurrences in patients with breathing disorders during sleep. These arousals can result from external stimuli or from an internal stimulus, interpreted as having a sleep protective role. This study will focus on automatic detection of arousals, especially those caused by sleep-related breathing disorders (apnea, hypopnea), based on all-night measurements of several subjects classified as having upper airway resistance syndrome (UARS). The detection algorithm follows the scoring rules defined by the American Sleep Disorder Association (ASDA) and the detection algorithm is based on frequency analysis of about 8 hours of sleep on a single EEG channel, and using other available channels to validate the results. We concluded that the automatic detector has a detection rate of about 70% which is comparable to reported visual analysis inter scorer agreement rates.