Breast Density Classification: a comparison of Different Classification Schemes

Authors

Abstract

Although the density of breasts is not a disease, it can lead to indiscrimination of some diseases, since it is hard for specialists to discern tumours in dense tissues. In this paper, based on tissue density, a mammogram is classified in four ordinal classes. To do this, an ordinal classification method is implemented with two different machine learning methods, SVM and Neural Network. Experimental results indicate that by using preprocessing followed by Support vector regression one attain better results than the other implemented methods.

Keywords

-breast density,Support vector regression,classification

Conference

StudECE 2013