Neural Systems Approach for Ammography Finding by Utilizing Wavelet Features

We present an application of artificial neural networks to mammographic images, aimed at improving early detection of sensitivity to breast cancer. The proposed application consists of two main steps: a pre-treatment step whose role is to extract the characteristics of the available mammographic images using the wavelet and co-occurrence (GLCM) matrices approach and a classification step based on an artificial neural network that uses these characteristics as input vectors for its training algorithm. The output of the training phase of this model is a categorization of the pre-treated images into three main groups: normal, benign and malignant. After the training phase, the network can be used in order to label new and unseen images as one of these three types.