Author : Renuka Mohan Khati and Rajesh Ingle

Epilepsy is one of the most common neurological disorders. Approximately 50 million of the population is affected by epileptic seizures worldwide. The diagnosis of epileptic seizure is a time consuming process. Moreover 75% of the epilepsy patients cannot even afford the expenses of the treatment of the epilepsy. An automatic approach is proposed for the detection of the epileptic seizures using the EEG signals. The detection of epilepsy requires potential features to be extracted from the EEG signal. The proposed approach will help the neurologists to detect the epileptic seizures with higher accuracy. In our approach we have used the power of convolutional neural network to extract robust features and have achieved an accuracy of 90%.

Keywords: Epilepsy; deep learning, convolutional neural network; Preictal; ictal