Tumor classification using radar target signatures

The use of Ultra Wideband (UWB) radar to detect early-stage breast cancer has been extensively investigated. The basis for this imaging modality is the significant dielectric contrast between normal and cancerous breast tissue at microwave frequencies. However, based on the recently-established dielectric similarities between malignant, benign and fibroglandular tissue within the breast, differentiating between these types of tissues in microwave images may be problematic. Therefore, it is important to investigate alternative methods to analyse and classify dielectric scatterers within the breast. One such approach is to classify scatterers based on their radar target signature, which carries information about scatterer size and shape. Benign tumors tend to have smooth surfaces and are compact and oval in shape. Conversely, malignant tumors tend to have rough and complex surfaces with spicules or microlobules. These properties can significantly influence the radar target signature, potentially allowing for dielectric scatterer classification. This paper presents a method to model the growth pattern of benign and malignant tumors, based on the use of Gaussian Random Spheres (ranging between smooth, macrolobulated, microlobulated and spiculated shapes), while classification algorithms that attempt to define the nature of tumors based on radar target signatures are also examined.