Detecting EEG bursts after hypoxic-ischemic injury using energy operators

During recovery following episodes of hypoxic-ischemic (HI) injury in the neonate, the electroencephalogram (EEG) recovers with sporadic, fluctuating energy discharges known as `bursts' and periods of electrical silence (`burst suppression'). Prior to the resumption of normal activity, the individual pattern of bursting may hold important diagnostic information regarding neurological outcome. Detection and characterization of the bursts seems to be an important factor in understanding the dynamics of the recovering EEG. The Teager Energy Algorithm (TEA) is a new method to describe abrupt EEG energy changes. Prior to the resumption of continuous EEG we coupled the use of the TEA and sequential detection to describe the start and stop time of bursts. Employing a dominant frequency model, the TEA provides distortion free reproduction of signal energy without the need for filtering high or sum frequency components portions. In an animal model of neonatal HI injury, we showed that TEA provides efficient detection of burst and burst suppression episodes. Burst counts might provide indicators of neurological and behavioral outcomes.