Thanks to the recent progresses in healthcare management, most patients survive after acute brain injury and the society must face several issues concerning the growing burden of disability and ethical questions related to withdrawal of life sustaining treatment in absence of certain prognostic markers. Describing more systematically coma awakening using a multimodal battery would help to refine consciousness re-emergence and to delineate biomarkers suitable for prognostication, which is exactly the aim of our project. Thus, we will perform a systematic longitudinal recording of coma awakening using such battery in Intensive Care Unit patients. We hypothesise that the current vision of successive steps of coma recovery (i.e. from a clinical class to another) is a partial view as it does not take into account the cyclic nature of wakefulness fluctuations. We propose that the best correlate for the natural history of coma recovery consists in a progressive shift from the loss of physiological rhythms toward a circadian phase of awakening clues allowing a rich cognitive content. Thus, we will monitor the circadian rhythmicity of biological markers and pro-wake system functionality (by biological dosages and EEG recording) during consecutive 24h once a week during the first month. We will also evaluate the lesional pattern in MRI and the expression of circadian clock genes. We expect that circadian rhythms may be present in some patients at the early phase since the very first days and could be used as biomarkers to anticipate coma awakening and functional outcome at 6 months. Thus, our objective is to identify patients who keep a sufficient functional reserve for a fast recovery after severe brain injuries and patients with an acute disruption of rhythms who might need complex brain plasticity to present a progressive improvement (if any) and to mechanistically explain these distinct patterns of recovery by the lesional and the genetic background.