Automatic detection of epileptiform activity using median filter and expert rule base

We present a new method to detect epileptiform activity based on median filter and expert rule base. Firstly, we reject the basic artifacts by applying a median filter to the original EEG wave. In this way, we are able to get the higher detection rate in real time which shows a specific noise pattern like EMG. Secondly, after converting the EEG wave into segment and sequence, we find a epileptiform activity by using the Gotman method. Then with expert rule system, we decrease detection error and validate epileptiform activity. The results show better efficiency of artifact rejection. Also, the detection algorithm makes many detection results because of lowering the threshold to find real epileptiform wave. In clinical results, our expert rule system is capable of rejecting artifacts commonly found in EEG recordings. By using the expert rule system, detection rate is found to be higher and shows the capability of real time detection