A method for the adaptive design of power spectral parameters using artificial neural networks and its application in the EEG-classification during brain ischemia

An adaptive method of parameter and sample design was proposed. It includes properties of discriminatory analysis and sample size design. By means of an included artificial neural network complex parameter patterns are also classified. The interesting signal parameters are estimated in a preprocessing unit. The optimization of the preprocessing and the artificial neural network was done in a coordinated way. The function of the proposed method was investigated using simulated and measured EEG data. First results concerning the classification of EEG power spectral parameters during cerebral ischemia are shown