Epilepsy is one of the most common neurological disorders.It is characterized by recurrent and transient epileptic seizures,which are seriously harmful for the human health both in physiology and psychology.The traditional visual examination of long-term EEGs by an experienced neurologist is a time consuming and subjective process.Therefore,it has been an interesting research area to study the automatic epileptic seizure detection using EEGs in recent years.In order to realize it successfully,how to design an appropriate feature extraction method becomes an important issue.This paper first proposes a novel fusion feature extraction method,and then combines extreme learning machine(ELM)and support vector machine(SVM)to complete the epileptic seizure detection automatically.The main contents are presented as follows:Chapter 1 systematically introduces the background,basic procedure and development of automatic seizure detection using EEGs.Chapter 2 mainly introduces the knowledge of EEG,and presents several most common used feature extraction methods and classifiers in the study of automatic epileptic seizures detection.Chapter 3 first proposes an improved Hjorth-parameter feature(IHP-F)and the second-order-difference sample entropy(SOD-SE)respectively.Then a novel fusion feature extraction method is designed by combining IHP-F and SOD-SE.Chapter 4 verifies the performance of the proposed novel fusion feature on Bonn EEG database.Experimental results show that the proposed feature extraction method is feasible and efficient in the automatic epileptic seizure detection. |