An Automatic Epileptic Seizure Recognition Using Two-Dimensional Convolutional Neural Network and Scalp EEG Signals

Epilepsy affects many people around the world. Experts usually detect epileptic seizures manually, but an intelligent system is required as it is a tedious and time-taking process and may cause human errors. In recent years, deep learning has been used in various medical applications, but still, It has not reached its maximum development potential. This paper presents a simple deep learning-based model. ElectroEncephaloGraphy (EEG) signals are plotted and directly fed into a convolutional neural network (CNN) model as input data. Through a CNN in a binary classification problem, the model learns to distinct seizures from non-seizures. The proposed method is superior and achieved 100% accuracy on the small sample of the Bonn University scalp EEG dataset.