The investigators want to validate the Sleep Diversity Index as a diagnosis tool of sleep fragmentation on 405 patients who will perform a polysomnography under spontaneous breathing.

Sleep fragmentation is one of the main problems of sleep disorders especially in respiratory failure. During this last decade, several indicators of sleep fragmentation were introduced but none of them seems to take into account the succession of sleep stages. To tackle with this problem, The investigators used the Shannon entropy index for modelling sleep stages diversity from an hypnogram and for assessing more accurately sleep quality by quantifying its fragmentation. The investigators developed a new tool we called Sleep Diversity Index (SDI).

In a retrospective study, the investigators calculated Sleep diversity index for 55 healthy subjects and 56 patients with suspected sleep apnea syndrom. Using the receiver operating characteristic (ROC) curves the threshold of fragmentation was determined at 20%. If SDI is lower than 20%, then sleep is considered as non-fragmented and otherwise, sleep is considered as fragmented.

To validate our SDI it is necessary to compare this diagnosis tool to a "gold standard" based on the clinician diagnosis and a mathematical diagnosis index created from objective criteria of polysomnography. The investigators propose a multicenter study to validate the SDI as a diagnosis tool of sleep fragmentation.

Inclusion Criteria:

• Spontaneous polysomnography

Exclusion Criteria:

• Insomnia
• Patients with psychoactive drug having an important effect on sleep

Patients who performed a spontaneous breathing polysomnography