Study And Application Of Automatic Detection Method For EEG

It is necessary to do EEG check-up for many patients with brain disease and neural system disease, for example brain trauma and brain tumor. It is reliable for epilepsy diagnosis. But EEG has many quantity of data generally, and its identification has many difficulties. Only EEG doctors have ability to identify. They also make different decision because of disturbance from artifacts. In order to raise efficency of diagnosis, shorten diagnosis'time, lower patient's pain, it is necessary to realize automation and intelligentization of detection for EEG. Brain wave is a kind of bioelectricity signal. Its charactered is that non-stationarity and low signal-noise rate. Developped methods are simulation method, expert system method, artificial neural network method, wavelet analysis method and non-linear dynamics method. But it is not enuough to use one method. The dissertation adopts many methods: DSP, wavelet analysis and neural network. They filter, resolve and identify brain wave in sequence. This makes better contribution to detection of epileptic wave. To realize automatic detection for EEG, the dissertation first designs two filters based on DSP technology. One is simple FIR filter, the other is lattice filter. They filter 50Hz artifacts and complicated artifacts separately. Then the dissertation utilizes wavelet analysis to resolve brain wave with different scales. It uses neural network as a tool for identification based on the principle that artifacts and epileptic wave must be different in different scales. esolved brain wave is input into neural network for identification. The output results in the identification of epileptic wave. The dissertation consists of six chapters. In the chaper 1, the dissertation preface, describes the background, current state, and the working of the research. In the chapter 2, it introduces basic theory of EEG's origin, illuminates composition and characteristic of brain wave, and origin and characteristic of artifacts. In the chapter 3, it presents filter designed in DSP and theory of wavelet analysis. In the chapter 4, it introduces artificial neural network, which is used to realize identification of epileptic wave. In iithe chapter 5, it provides system design scheme and implement process. In the last chapter of the paper, the contents of the paper are summarized and the next task are arranged.