Efficient Seizure Prediction from Images of EEG Signals Using Convolutional Neural Network

Epileptic seizures are abnormal electrical activities in the brains of epilepsy patients. Seizure causes life threats due to sudden unconsciousness. An efficient seizure prediction technique can improve the lifestyle of epilepsy patients. In this article, raw EEG data are converted into images of EEG signals. The Convolutional neural network (CNN) is the most commonly used deep learning technique for efficient feature extraction from images. Hence, CNN is applied here for automatic feature extraction from images of EEG signals. CNN is also used to classify different states of epilepsy patients to predict seizures in advance. The achieved classification accuracy, sensitivity, and specificity are 0.9994, 0.9783, and 0.9083, respectively. The proposed method uses only six channels of EEG signals, which will be applicable for designing lightweight devices with less power consumption to run long periods in every recharge. The results indicate that the proposed method is one of the best among the state-of-the-art works.