Frequency Model Filtering for Microwave Imaging in Breast Cancer Applications

Abstract

In 2020, over 27,000 women were diagnosed with and treated for breast cancer in Canada.  There are significant limitations with current imaging modalities being used in tracking recovery from treatment, such as ionizing radiation in x-ray mammography and accessibility for magnetic resonance imaging (MRI).  Microwave imaging has shown to overcome these limitations with remarkable ability to distinguish between healthy and non-healthy tissue.  The microwave imaging transmission system (MITS) developed at the University of Calgary can improve its results with filtering multipath data from the acquired frequency data.  An interpolation modelling method is proposed to adapt with different variations in the microwave signals using polynomials of orders up to 15. The polynomial models show representation of the dominant signal’s shape and width with high statistical significance in the R squared and F- tests; therefore, providing reliable filtering parameters to create the microwave images.