Abstract
Ultrawideband (UWB) microwave radar is a promising alternative breast screening method. Previous research has focused on the imaging and classification of early-stage breast cancer from backscattered microwave signals. The heterogeneous composition of breast tissue, prevalent among younger females, inhibits UWB scanning technologies to effectively localize cancerous regions within the breast. Rather than using UWB radar to image the breast or classify between types of cancer, the method proposed in this paper is to simply detect the presence of a tumor using a Support Vector Machine (SVM) based UWB cancer detection system. The SVM cancer detection system is evaluated using dielectrically realistic numerical breast models, and the performance of the detection system is compared to a Linear Discriminant Analysis (LDA) classification method.