Classification of magnetic resonance (MR) brain images using energy coefficient and neural network

Brain tumor is an abnormal growth of cells in the brain . Magnetic resonance imaging ( MRI ) is an advanced diagnostic tools that enable us to visualize anatomical details more clearly so superior in detecting abnormalities in the soft tissues of the brain . On this study , a data classification system built Magnetic resonance imaging ( MRI ) of the brain by using energy coefficients and neural network . On the results of brain MRI performed three stages of the process , namely feature extraction , feature reduction , and classification . Results in the form of feature extraction vector detail coefficients horizontal , vertical , diagonal and vector approximation of wavelet decomposition of each level . Feature is the result of the reduction of energy in the form of vector approximation coefficients of each wavelet decomposition level . In the process of neural network classification method is used to classify the types of normal brain disease , Alzheimer's disease , glioma and carcinoma . Percentage success rate of recognition obtained brain MRI features by 95 % by using 10 energy coefficient , learning rate of 0.4, the activation function is a function logsig the hidden layer and the output layer and the process stops at the epoch to 515