Implementation of classification system for Brain Cancer using backpropagation network and MRI

Brain Cancer is a fatal disease. Human analysis on medical images is a difficult task due to very minute variations. This paper presents extension in Computer Aided Diagnosis for early prediction of Brain Cancer Class using texture features and neuro classification logic and backpropagation artificial neural network. The work involves extraction of texture features from the given brain MRI sample using discrete wavelet transform and morphological operation followed by neuroclassification. Astrocytoma type of Brain Cancer is only considered for the study. Database is constructed from the case studies available on the internet. This paper is the extension of our previous paper based on locating tumor and extracting features from Brain Cancer affected MRI. The various backpropagation training algorithms are finally compared in order to build up efficient classification system.