Classification of suspicious regions within ultrawideband radar images of the breast

Ultrawideband (UWB) radar is one of the most promising emerging technologies for breast imaging. Several algorithms have already been developed which exploit dielectric contrasts between normal and malignant tissue. These algorithms have been tested on anatomically accurate models of the breast. However, the recently established similarities in dielectric properties between benign/fibroglandular and malignant tissue within the breast contribute to the occurrence of dasiafalse positivepsila results in UWB images. To mitigate the presence of these dasiafalse positivepsila results, we must investigate methods to further classify tumours based on their shape, size and texture. In this paper we review the dielectric properties of normal and malignant breast tissue and existing image algorithms. Finally, we examine methods to effectively differentiate benign/fibroglandular and malignant tissue based on shape, size and texture.