Epileptic seizure detection using orthogonal local preserving projection and k-nearest neighbors classifier
In present decades, automatic epileptic seizure detection is a growing research area in the field of medical diagnosis. Electroencephalogram (EEG) is an emerging medical device that effectively monitors the epileptic seizures. In this paper, epileptic seizure detection was assessed by using Bonn University dataset. After collecting the EEG signals, Hilbert Vibration Decomposition (HVD) and hybrid features (combination of entropy and statistical features) were applied to extract the features. Then, Orthogonal Local Preserving Projection (OLPP) was utilized to lessen the number of random features by achieving a set of principal features. After obtaining the optimal feature values, a supervised classifier (K-Nearest Neighbors (KNN)) was used to classify the epileptic seizure classes; normal, ictal, and interictal. The experimental study illustrates that the proposed system effectively classifies the epileptic seizure classes in light of sensitivity, specificity, Negative Predictive Value (NPV), Positive Predictive Value (PPV) and accuracy. The proposed system improves the classification accuracy upto 1.5-2% related to the existing systems.