Likeness-based detection of sleep slow oscillations in normal and altered sleep conditions: application on low-density EEG recordings.

Sleep slow oscillation (SSO) is a common EEG pattern of spontaneous activity during nonrapid eye movement sleep. A new method for detecting SSOs is presented and compared to previous canonical methods. The main result of this research is that for the first time an extensive SSO analysis is applied to clinical EEG montages, based on low-density EEG recordings. The proposed method gives positive indications about its effectiveness also for altered sleep, extending the SSO analysis to extreme cases, and thus, opening a new front for investigating pathophysiological correlates of sleep disorders.