M.Sc. Tezi Görüntüleme

Student: Önder AYDEMİR
Supervisor: Assoc. Prof. Dr. Temel KAYIKÇIOĞLU
Department: Elektrik-Elektronik Müh.
Institution: Graduate School of Natural and Applied Sciences
University: Karadeniz Technical University, Turkey
Title of the Thesis: Feature Extraction For EEG Signals Towards Brain Computer Interface Applications
Level: M.Sc.
Acceptance Date: 26/6/2008
Number of Pages: 93
Registration Number: i1914

      Electroencephalography (EEG) signals are the low amplitude bioelectrical signals which are received from brain surface. Those signals’ peak to peak amplitude is 1-400 μV and frequency band situates between 0.5-100 Hz. Since EEG signals include much information about brain activities, in last years the research in this area has accelerated in the fields of medicine and engineering. In the field of medicine, those signals play an important role in diagnosing neurological diseases and in monitoring success of medical treatments selected. In the engineering field, the features extracted from EEG signals recorded during the mental and visual tasks provide valuable information brain computer interfacing. Brain Computer Interface (BCI) makes possible to people to use a computer, an electromechanical arm or variety of neuroprothesis without using their muscle systems, in other words without using their motor neurosystems. Nowadays, a couple of techniques, such as EEG, single cell recordings (SCR), functional magnetic resonance imaging (fMRI), local field potential (LFP), Near Infrared Spectroscopy (NIRS), electrocorticography (ECoG) and magnetoencefalography (MEG) are used for BCI implementations. Among these techniques, generally EEG is selected for BCI systems due to the fact that it is easy to apply and practically accessible.

The goal in area of EEG based BCI research is to develop a method which has higher classification rate and brain computer interfacing data rate than existing methods.

       In this thesis, it was studied on EEG dataset which was obtained under different mental and visual tasks used in literature. Feature extractions from that EEG dataset were analysed with various mathematical methods. According to discovered features classification was done by using support vector machines (SVM) and KNN classifier. All these studies actualized by considering the goal of to obtain higher classification rate and high brain computer interfacing data rate.

      Key Words: EEG, BCI, Feature Extraction, Classification