No abstract

Research has been conducted to support an automatic diagnosis system that will relieve clinicians of their weary work by detecting epileptic seizures. In this paper, we suggest a novel method to automatically identify epilepsy crises based on electroencephalogram (EEG) and electrocardiogram (ECG) signals. The work to detect epileptic seizures from EEG and ECG signals is carried out in three stages. In the first stage, simultaneous EEG and ECG recordings captured from 24 channels are segmented into 10-second periods (where 23 are the EEG signals and one is the ECG signal). In the second stage, the extraction of the parameters of each channel from the time domain and, finally, the classification of the EEG and ECG signals into epileptic seizure and normal have been done using ANN. Experiment analysis shows that using the ECG signal as extra information has a high capacity for classification.