TY - CONF ID - SisLab1577 UR - https://eprints.uet.vnu.edu.vn/eprints/id/eprint/1577/ A1 - Tran, Duc Tan A1 - Nguyen, Linh Trung Y1 - 2016/01/01/ N2 - An automatic detection system for epileptic spikes is used as an assistive tool for epileptic analysis and diagnosis using electroencephalogram (EEG). Our system consists of four stages: preprocessing, feature extraction classification and expert system. In this paper, we concern to the performance of 4th stage (expert system) to choose the best threshold for detecting epileptic spikes efficiently. The proposed method is verified by experimental data acquired from ten patients (for training) and four patients (for testing). TI - Investigating the performance of expert system in a multistage system for automatic detection of epileptic spikes SP - 1 M2 - Hanoi, Vietnam AV - public EP - 5 T2 - JCC: the 2nd International Joint Conference on Convergence ER -