Epileptic activity recognition in EEG recording

We apply Approximate Entropy (ApEn) algorithm in order to recognize epileptic activity in electroencephalogram recordings. ApEn is a recently developed statistical quantity for quantifying regularity and complexity. Our approach is illustrated regarding different types of epileptic activity. In all segments associated with epileptic activity analyzed here the complexity of the signal measured by ApEn drops abruptly. This fact can be useful for automatic recognition and detection of epileptic seizures.