BRIEF SUMMARY :

In the most severe cases of brain injuries, intensive care may allow patients with altered consciousness to survive despite a significant risk of heavy sequelae. Persistent impairments of consciousness are currently categorized according to behavior in three main neurological categories: comatose state, vegetative state (recently named unresponsive wakefulness syndrome) and minimally conscious state. Refining the diagnosis of internal state is a major goal to determine the abilities for an optimal recovery of cognitive deficit. Circadian rhythms are implicated in the regulation of sleep-wake cycles but also in cognitive functions. Their role is actually revaluated in the mechanisms of consciousness impairment. First, it is well known that cognitive performances partially depend on such rhythms as they are more elevated during the day and correlated to the hormonal secretion. In a prognostic point of view, fewer rhythmic perturbations during the initial resuscitation period (with reorganized sleep rhythms and the presence of paradoxical sleep) could be associated to a higher functional outcome.

However, this internal state of alertness could be highly variable during the day as it might be influenced by specific rhythms such as the circadian rhythm. Only a continuous assessment could help defining them properly.

Thus, investigators hypothesize that the circadian restauration, assessed in a dynamic perspective, is associated with the improvement of content and level of awareness. The main challenge of our study is to capture the long-term changes in the evolution of circadian and ultradian rhythms and to keep a part of the natural history of the clinical recovery of these patients.

To achieve this goal, the investigators plan to analyze during more than 2 days both neurophysiological rhythms (EEG) and behavioral rhythms of alertness ("Eyes" scale from of the Glasgow coma scale) in a dataset collected retrospectively from the population of patients continuously monitored by EEG for medical purposes (to identify seizures and prevent status epilepticus) in an intensive care unit of teaching hospital as far as acquisitions last more than 48h and present no prolonged epileptic discharges or artifacts leading to uninterpretable EGG.

Inclusion Criteria:

• Adult patient
• Admitted in ICU
• Altered level of consciousness not explained by a continuous sedation
• EEG during more than 48h consecutively

Exclusion Criteria:

• Prolonged epileptic discharges leading to uninterpretable EEG rhythms
• Prolonged artifacts leading to uninterpretable EEG trace

Comatose patient in ICU under intermittent behavioural evaluation and continuous EEG recording.