EEG waveform detection based on the search of distinctive line-segments

In this paper we describe a new EEG waveform detection method, based on the search of distinctive line-segments. The method is based on the assumption that EEG waveform morphology is characterized by a sequence of its distinctive line-segments and their structural features. In this method, the distinctive line-segments are first searched for, and the structural feature analysis is performed on the found distinctive line-segment sequences. The method does not need a priori signal conditioning. Experiments of detecting epileptic spikes are performed on for four different subjects.