Automatic recognition and quantification of interictal epileptic activity in the human scalp EEG

Abstract

An attempt was made at using a small computer to recognize and quantify interictal epileptic activity (spikes and sharp waves) in the human scalp EEG. To perform the automatic recognition, the EEG of each channel is broken down into half-waves. A half-wave is characterized by its duration and its amplitude relative to the background activity. A wave is characterized by the durations and amplitudes of its two component half-waves, by the second derivative at its apex measured relative to the background activity, and by the duration and amplitude of the following half-wave. Particular combinations of these parameters were found to characterize spikes and sharp waves and are used for their recognition and quantification. Specific methods are used for the rejection of spike-like or sharp wave-like wave forms such as eye blinks, muscle potentials and sharp alpha activity and were found to perform with a high level of reliability. Interchannel relationships are thoroughly examined to determine areas of maximal epileptogenicity. Sixteen channels can be analyzed in real time. Results are presented in a simple picture containing localizing and quantitative information. Specific questions regarding the time relationships of spikes in different channels can be asked interactively by the user. The system is of potential use in clinical electroencephalography.