International Journal of Research Studies in Science, Engineering and Technology Volume 2, Issue 7, July 2015, PP 1-12 ISSN 2349-4751 (Print) & ISSN 2349-476X (Online) International Journal of Research Studies in Science, Engineering and Technology [IJRSSET] 1 Assessment of EEG Signals Using Chaos Analysis Methods Muhittin Bayram Department of Electrical and Electronics Engineering Dicle University, Diyarbakir, Turkey Abstract: Electroencephalogram (EEG) signals carry information about the dynamics of the brain. A nonlinear method development is of great significance objective or goal because of the brain signals are nonlinear. Epilepsy is a neurological disorder which can be seen all over the world. It can be diagnosed by brain’s electrical activity. The determination of epileptic attacks or seizures by EEG signals is quite common in both clinical and research fields. During epileptic seizures, brain dynamics that make up the graph consists of abnormalities in EEG signals. Therefore, both time and frequency bands of the signals as well as need to go to review and determination of the pattern. Preictal and ictal EEG epochs were evaluated by wavelet-entropy and artificial neural networks (ANN) methods in this study. One hour EEG signals from different patients were used for wavelet-entropy method. One-hour epileptic EEG signals divided into two states (preictal and ictal) for this study. Then preictal end ictal states divided into 20 second segments. All of EEG segments have been separated into the standard subbands. Shannon entropies of the EEG subbands are calculated; and finally the feature vectors are classified with ANN. As a result, preictal and ictal EEG subbands’entropy values distinguish preictal and ictal segments from each other and the distinctiveness of gamma subbands’ entropy values are more robust. The composite system that was proposed using performance evaluation criteria showed a 97,5 % success rate in classification. Keywords: EEG, Epilepsy, Wavelet-entropy, ANN. 1. INTRODUCTION Electrical signals derived from dynamics of biological organs are used in many areas of medicine, diagnosis and treatment. EEG, ECG and other biological signals directly or indirectly provides information about the condition of the organ involved. Biological signals are processed by different signal processing methods such as time, frequency, time-frequency and phase. In the light of these new strategies are being developed in the medical field. In the context, EEG signals were examined by chaos analysis methods to obtain the information about the dynamics and the state of the brain. Also, considered will gain a new dimension to the pathological EEG signals. A negativity of the brain affects the brain and the body with synchronized between the brain and all of the organs in the body due to the interactive structure with the body. Similarly, a negativity of the body affects the body together with the brain. EEG signals carry information about the dynamics of the brain. The accuracy of this information as possible to understand of the development of methods that can be used is very important. A nonlinear method development is of great significance objective or goal because of the brain signals are nonlinear. Epilepsy is a neurological disorder which can be seen all over the world. It can be diagnosed by the brain’s electrical activity. The determination of epileptic attacks or seizures by EEG signals is quite common in both clinical and research fields. Because EEG signals are non-stationary signals, they must be examined with the nonlinear analysis methods. During epileptic seizures, brain dynamics that make up the graph consists of abnormalities in EEG signals. Because of the dynamics of these signals are due to several factors are not easy to put the required informations. Therefore, both time and frequency bands of the signals as well as need to go to the review and determination of the pattern. Identifying patterns of the signals depends on the extraction of the characteristics. Defining the degree of a property of an event is about separation in the event of sharp lines from others. Therefore, the incident event clearly must could be characterize the detected property. In this study, the dynamics of the brain contribute to the development of detection and early diagnosis systems and EEG signals in order to achieve the results form the basis for the determination of