“Relevance vector machine” consciousness classifier applied to cerebral metabolism of vegetative and locked-in patients

►Vegetative State (VS) patients remain unconscious after awaking from their coma.►Locked-in Syndrom (LIS) patients recover full conscious after awaking from their coma but remain unable to communicate.►Bedsite evaluation of LIS and VS patients is difficult and prone to error.►Using FDG-PET images, we train a "relevance vector machine classifier" (RVMc) to distinguish unconscious VS patients from healthy conscious controls.►Applying the trained RVMc on LIS and VS FDG-PET images then automatically rates the LIS patients as conscious and VS as unconscious.