Classification of Brain Tumours from The MR Images Using Neural Network and Central Moments

MRI can detect a wide range of brain conditions, including swelling, tumours, cysts, bleeding, structural abnormalities, infections, inflammatory conditions, and blood vessel problems. The main goal was to generate the distribution of every zone selected by moving the window of size 16 by 16 pixels on the MR picture of brain, resulting in 64 histograms and each collected histogram will be evaluated as a series and for this the central moments of order one, two, and three will be calculated. A multilayer perceptron performs the classification i.e., brain tumour classification using MR images. Database used was made up of the collection of MR pictures of the brain that have been mixed with various types of brain tumours which belonged to unique people. The 3 steps which comprise the proposed system are given namely, pre-processing in this step the size of MR brain pictures where normalized and converted, feature extraction where histogram’s zone that are obtained after sliding a 16 by 16-pixel window on image and the order one, two and three of central moments are calculated, as well as the classification step carried out with the help of a perceptron with multiple layers