High frequency oscillations in intracranial recordings of epileptic patients

The seizure detection in Electroencephalogram (EEG) of epileptic patients is complex as the events can be rare and unpredictable. This paper describes a relatively new approach for the identification of epileptic seizures in EEG using High Frequency Oscillations (HFOs). The method is based on the calculation of band power in the high frequency band (80-200 Hz) after dividing EEG in to epochs of one second. The values obtained are compared with the pre-set threshold and a grouping scheme is followed to ignore isolated detections. The method is able to detect all (but one) the seizures in the tested patients. The false alarm rate is observed to vary from 0 to 0.5 per hour. The results strongly suggest that HFO can be used for automatic detection with a high sensitivity and low false alarm rate.