Automatic recognition and characterization of epileptiform discharges in the human EEG.

Methods proposed for the automatic identification and quantification of epileptiform EEG activity are reviewed, and the potential role of this technology in clinical electroencephalography is assessed. Techniques developed for the detection of spikes, sharp waves, and spike and wave complexes are described. Emphasis is placed on the problems associated with artifact rejection and the need for establishing context-based decision-making processes.