Analysis for Automatic Detection of Epileptic Seizure from EEG signals

During longterm EEG monitoring of epileptic patients, automatic detection methods could be of great assistance because they save a lot of time. The work was develop in cooperation with Micromed Spa with the aim of evaluate and compare the performance of two seizure detection algorithm:one using wavelet based features and one based on AR parameters. The Artificial Neural Network were used as classification method. The results show a better reliability for the ANN having in input AR parameters